{"name":"Who Cried AGI?","url":"https://whocriedagi.com/claims","updatedAt":"2026-10-23","description":"Sourced public statements, not a consensus forecast or a determination that AGI exists. Statement dates and archive revision dates are separate.","license":"https://creativecommons.org/licenses/by/4.0/","licenseScope":"CC BY 4.0 applies only to original annotations and the compilation, to the extent Who Cried AGI? holds rights. Third-party quotations and source material are excluded and retain their existing rights.","licenseDetails":"https://whocriedagi.com/about#dataset-license","records":[{"id":"hassabis-2026-wsj-two-three-years","url":"https://whocriedagi.com/claims/hassabis-2026-wsj-two-three-years","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2026-10-21","datePrecision":"day","quote":"He expects AGI within two to three years, roughly 2028–2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"AI Weekly and Trending Topics report that a Wall Street Journal profile shortened Hassabis's outlook from three to five years earlier in 2026 to two to three years. AI Weekly explicitly identifies the timing as the Journal's paraphrase, not a direct quotation. This entry uses the profile's reported publication date; the interview date is unknown. The paywalled original was not independently inspected.","source":"https://aiweekly.co/alerts/hassabis-trims-agi-window-to-two-to-three-years-in-wsj-profile","sourceLabel":"AI Weekly / The Wall Street Journal","evidence":"AI Weekly's October 21 report identifies Ben Cohen's Wall Street Journal profile, its publication date and the paraphrased two-to-three-year forecast. Trending Topics independently reports the same window on October 21. Both secondary reports were checked; direct access to the WSJ article returned HTTP 401.","target":"Two to three years: roughly 2028–2029","expectedArrival":{"sourceDate":"2026-10-21","sourceDatePrecision":"day","kind":"range","startYear":2028,"endYear":2029,"classification":"ahead","approximate":true,"interpretation":"Two to three years from the October 2026 profile gives roughly 2028–2029. The forecast is a reported paraphrase and the interview date is unknown, so no exact arrival month, day or probability is inferred.","source":"https://aiweekly.co/alerts/hassabis-trims-agi-window-to-two-to-three-years-in-wsj-profile"},"createdAt":"2026-10-22","updatedAt":"2026-10-22"},{"id":"benaich-2026-state-of-ai-2027","url":"https://whocriedagi.com/claims/benaich-2026-state-of-ai-2027","name":"Nathan Benaich","role":"General Partner","company":"Air Street Capital","date":"2026-10-08","datePrecision":"day","quote":"AGI 2027.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The State of AI Report 2026 lists this forecast among nine predictions for the next twelve months. The prediction does not specify an AGI definition, a model, or a probability.","source":"https://www.stateof.ai/State-of-AI-Report-2026.pdf#page=239","sourceLabel":"State of AI Report 2026 · page 239","evidence":"Primary PDF visually checked: the cover names Nathan Benaich and dates publication October 8, 2026; PDF page 239 (printed slide 238) contains the exact prediction. The author biography identifies him as General Partner of Air Street Capital.","target":"2027","expectedArrival":{"sourceDate":"2026-10-08","sourceDatePrecision":"day","kind":"range","startYear":2027,"endYear":2027,"classification":"ahead","interpretation":"The report states AGI 2027 among its predictions for the next twelve months. The stored window preserves the explicit calendar year without inventing a precise arrival day or a numerical probability.","source":"https://www.stateof.ai/State-of-AI-Report-2026.pdf#page=239"},"createdAt":"2026-10-08","updatedAt":"2026-10-08"},{"id":"altman-2026-politico-decoded-agi","url":"https://whocriedagi.com/claims/altman-2026-politico-decoded-agi","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2026-10-05","datePrecision":"day","quote":"I think a lot of people would say AGI happened or is close to happening.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"Asked about AI terminology in a POLITICO Decoded interview with Brendan Bordelon, Altman described what others might consider AGI's arrival while leaving open whether it was only near. He supplied no arrival deadline or model-specific claim. The wording comes from a third-party transcript; the original recording has not been independently verified.","source":"https://ceointerviews.ai/interview/2985049/","sourceLabel":"POLITICO Decoded / CEOInterviews.AI transcript · 3:44","evidence":"CEOInterviews.AI attributes the excerpt at 3:44 to Sam Altman and lists October 5, 2026 as the interview's publication date. That listed publication date is used here; the recording date is not established. The transcript links to POLITICO's YouTube video at 224 seconds. The original video could not be independently checked, so this is transcript-based evidence rather than primary-recording verification.","createdAt":null,"updatedAt":"2026-10-23","revisionTrackingStartedAt":"2026-10-22"},{"id":"schulman-2026-ford-pro-agi","url":"https://whocriedagi.com/claims/schulman-2026-ford-pro-agi","name":"Dan Schulman","role":"CEO","company":"Verizon","date":"2026-09-30","datePrecision":"day","quote":"I think we get to some form of AGI in the next six to 18 months.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Speaking with Fortune's Diane Brady at Ford Pro Accelerate in Detroit, Schulman forecast some form of AGI in six to eighteen months, roughly March 2027–March 2028. He did not supply a specific AGI test or name a qualifying model. His separate predictions about quantum computing and humanoid robotics are not AGI deadlines.","source":"https://fortune.com/2026/09/30/this-revolution-is-happening-in-years-not-decades-verizon-ceo-dan-schulman-sees-agi-within-18-months-and-a-century-of-progress-in-a-decade/","sourceLabel":"Fortune / Ford Pro Accelerate","evidence":"Nick Lichtenberg's September 30, 2026 event report quotes the six-to-eighteen-month forecast and identifies Schulman's conversation with Diane Brady at Ford Pro Accelerate that day. The published report was checked; the event recording was not independently reviewed.","target":"Six to eighteen months: roughly March 2027–March 2028","expectedArrival":{"sourceDate":"2026-09-30","sourceDatePrecision":"day","kind":"range","startYear":2027,"endYear":2028,"deadlineMonth":3,"classification":"ahead","approximate":true,"interpretation":"Six to eighteen months from September 30, 2026 gives roughly March 2027–March 2028. The calendar uses 2027–2028 and retains March as the endpoint month without inventing an exact arrival day. Schulman's qualifier 'some form of AGI' is preserved; no numerical probability or precise capability threshold is supplied.","source":"https://fortune.com/2026/09/30/this-revolution-is-happening-in-years-not-decades-verizon-ceo-dan-schulman-sees-agi-within-18-months-and-a-century-of-progress-in-a-decade/"},"createdAt":"2026-10-22","updatedAt":"2026-10-22"},{"id":"social-x-2100229706641539248","url":"https://whocriedagi.com/claims/social-x-2100229706641539248","name":"Shane Legg","role":"Cofounder and chief AGI scientist","company":"Google DeepMind","date":"2026-09-16","datePrecision":"day","quote":"AGI is approaching, making deeper study of its technical and social consequences urgent.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Legg announces the DeepMind Institute. The post supplies no date for AGI’s arrival.","source":"https://x.com/ShaneLegg/status/2100229706641539248","sourceLabel":"Shane Legg on X","evidence":"Author and statement text verified in X’s official public oEmbed response; UTC date computed from the original Snowflake post ID and checked against the embed’s date.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2097148694995103889","url":"https://whocriedagi.com/claims/tweet-sweep-x-2097148694995103889","name":"Aravind","role":"Commentator · @aravind","company":"","date":"2026-09-08","datePrecision":"day","quote":"Aravind reiterates a 2028 AGI forecast and places AI exceeding specialist humans, which he calls ASI, in the 2040s.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"The @aravind account distinguishes something he would call AGI in 2028 from systems better than specialist humans in the 2040s. Attribution is to this account, not to Perplexity CEO Aravind Srinivas. The plotted target is the AGI year; the separate ASI decade is retained here.","source":"https://x.com/aravind/status/2097148694995103889","sourceLabel":"Aravind on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2026-09-08","kind":"around","targetYear":2028,"interpretation":"AGI in 2028; a separate ASI horizon is the 2040s.","note":"This record contains two thresholds. The displayed AGI target does not represent the ASI decade."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"huang-astra-2026","url":"https://whocriedagi.com/claims/huang-astra-2026","name":"Jensen Huang","role":"Co-founder and CEO","company":"NVIDIA","date":"2026-09-06","datePrecision":"day","quote":"AGI has arrived.","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-6 Astra"],"context":"Huang congratulated OpenAI after Astra's release.","source":"https://x.com/JensenHuang/status/2096700264569090384","sourceLabel":"Jensen Huang on X","evidence":"User-supplied post; date and wording corroborated by Business Insider's September 7 report.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2096445087505055891","url":"https://whocriedagi.com/claims/social-followup-x-2096445087505055891","name":"Chase Lochmiller","role":"Co-founder and CEO","company":"Crusoe","date":"2026-09-06","datePrecision":"day","quote":"Congratulations to our friends at @OpenAI! I guess this makes Abilene the birthplace of AGI!","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-6 Astra"],"context":"Lochmiller congratulates OpenAI and describes the Abilene training site as AGI’s birthplace. The attached link points to OpenAI’s GPT-6 Astra benchmark announcement; this is his own claim, preceding Huang’s reply.","source":"https://x.com/ChaseLochmiller/status/2096445087505055891","sourceLabel":"Chase Lochmiller on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2096728551269839076","url":"https://whocriedagi.com/claims/social-followup-x-2096728551269839076","name":"Mark Chen","role":"Chief research officer","company":"OpenAI","date":"2026-09-06","datePrecision":"day","quote":"Agree with @JensenHuang: we’re entering the AGI era.","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-6 Astra"],"context":"Chen explicitly agrees with Jensen Huang’s Astra-era AGI declaration and calls for capable AI monitors and alignment with humanity. The post endorses the arrival claim while emphasizing oversight.","source":"https://x.com/markchen90/status/2096728551269839076","sourceLabel":"Mark Chen on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"brockman-astra-2026","url":"https://whocriedagi.com/claims/brockman-astra-2026","name":"Greg Brockman","role":"Co-founder and President","company":"OpenAI","date":"2026-09-03","datePrecision":"day","quote":"For me personally, I do think we're there.","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-6 Astra"],"context":"His personal assessment of Astra; he invited others to decide whether it met their definition.","source":"https://www.businessinsider.com/nvidia-jensen-huang-agi-openai-astra-ai-2026-9","sourceLabel":"Business Insider","evidence":"September 7 report quotes Thursday's briefing; September 3 reporting by Axios and The Washington Post independently corroborates the claim.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2095607046129463577","url":"https://whocriedagi.com/claims/social-followup-x-2095607046129463577","name":"François Chollet","role":"Co-founder","company":"Ndea / ARC Prize","date":"2026-09-03","datePrecision":"day","quote":"Sooner, given progress is happening faster than I expected.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Responding to a direct question about his earlier “AGI ~2030” forecast, Chollet says the timeline is sooner after Astra’s results. The question and answer were both verified through official X embeds. He supplies no replacement year, so the archive does not invent a narrower window. This is a forecast revision, not a claim that Astra is AGI.","source":"https://x.com/fchollet/status/2095607046129463577","sourceLabel":"François Chollet on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2026-09-03","kind":"unspecified","interpretation":"Sooner than his previous estimate of around 2030, without a new date.","note":"Do not turn a revision relative to an approximate year into an exact 2029 deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"irving-2026-80000-superintelligence","url":"https://whocriedagi.com/claims/irving-2026-80000-superintelligence","name":"Geoffrey Irving","role":"Cofounder and chief scientist","company":"Resolution","date":"2026-08-11","datePrecision":"day","quote":"His modal expectation was full superintelligence within two to three years, while acknowledging it might take 10–20.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"He identified generalization from verifiable tasks to judgment and long-term planning as a key uncertainty. Recorded June 29, 2026; date is public release.","source":"https://80000hours.org/podcast/episodes/geoffrey-irving-superintelligence-alignment-theory/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: Why Geoffrey expects superintelligence in 2–3 years, 00:28:05.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"ord-2026-80000-timelines","url":"https://whocriedagi.com/claims/ord-2026-80000-timelines","name":"Toby Ord","role":"Senior researcher","company":"University of Oxford, AI Governance Initiative","date":"2026-08-06","datePrecision":"day","quote":"His median date for transformative AI was 2038, but his uncertainty extended from a few years to more than a century.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His threshold included doubling scientific progress and potentially taking over the world. Recorded July 2, 2026; he advocated preparing for both short and long timelines.","source":"https://80000hours.org/podcast/episodes/toby-ord-recursive-self-improvement-agi-timelines/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript and attributed highlights: timelines discussion, 01:09:28 and 01:45:00.","target":"2038 median; broad uncertainty","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"shah-2026-80000-explosion","url":"https://whocriedagi.com/claims/shah-2026-80000-explosion","name":"Rohin Shah","role":"Head of AGI safety and alignment","company":"Google DeepMind","date":"2026-06-02","datePrecision":"day","quote":"The transition from the first automated AI researchers to an intelligence explosion would likely take years rather than months.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"He expected early automated researchers to be expensive and inefficient and pushed back on imminent-explosion arguments. Recorded December 4, 2025; date is public release.","source":"https://80000hours.org/podcast/episodes/rohin-shah-google-deepmind-agi-safety/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: Why Rohin doubts the intelligence explosion is imminent, 01:52:44.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"andreessen-2026-rogan-agi","url":"https://whocriedagi.com/claims/andreessen-2026-rogan-agi","name":"Marc Andreessen","role":"Co-founder and General Partner","company":"Andreessen Horowitz","date":"2026-05-19","datePrecision":"day","quote":"And I actually think we crossed that about 3 months ago.","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-5.5","Claude 4.6","Gemini 3.0","Grok 4.3"],"context":"He defined AGI as being as smart as a person and cited his experience getting expert-quality answers. These model versions are the names he used, not an independent assessment of their capabilities.","source":"https://a16z.com/podcast/marc-andreessen-on-ai-california-and-the-future-of-america-joe-rogan/","sourceLabel":"The Joe Rogan Experience / a16z","evidence":"JRE #2501's official podcast listing dates publication May 19; a16z republished May 20. Happy Scribe and Podscripts transcripts both attribute all four versions to Andreessen around 1:45.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2046701483970949601","url":"https://whocriedagi.com/claims/tweet-sweep-x-2046701483970949601","name":"Robert Levy","role":"Commentator · @rplevy","company":"","date":"2026-04-21","datePrecision":"day","quote":"Levy identifies ChatGPT’s 2022 release as AGI’s arrival, correcting his own earlier claim that the release was in 2023.","paraphrase":true,"kind":"qualified","topic":"agi","models":["ChatGPT"],"context":"Levy treats the initial release of ChatGPT as the release of AGI and corrects the year from 2023 to 2022. This is his interpretation of AGI, published in April 2026; the record is not backdated to 2022.","source":"https://x.com/rplevy/status/2046701483970949601","sourceLabel":"Robert Levy on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2040922415551959338","url":"https://whocriedagi.com/claims/social-followup-x-2040922415551959338","name":"Marc Andreessen","role":"Co-founder","company":"Andreessen Horowitz","date":"2026-04-05","datePrecision":"day","quote":"I'm calling it. AGI is already here – it's just not evenly distributed yet.","paraphrase":false,"kind":"arrived","topic":"agi","models":[],"context":"Andreessen makes an explicit arrival claim without naming a model or defining AGI. This April social post is distinct from his later May podcast interview already in the archive.","source":"https://x.com/pmarca/status/2040922415551959338","sourceLabel":"Marc Andreessen on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2037360219303690628","url":"https://whocriedagi.com/claims/tweet-sweep-x-2037360219303690628","name":"Eli Lifland","role":"Forecaster","company":"","date":"2026-03-27","datePrecision":"day","quote":"I personally have an ASI median of ~2035","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"Lifland gives an approximate median arrival year for ASI. The separate extinction probability and conditional timing in the same post are different forecasts. The public embed truncates the end of the post, but the ASI median is fully visible.","source":"https://x.com/eli_lifland/status/2037360219303690628","sourceLabel":"Eli Lifland on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2026-03-27","kind":"around","targetYear":2035,"approximate":true,"probability":{"value":0.5},"interpretation":"Approximate median ASI arrival year of 2035."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"huang-lex-2026","url":"https://whocriedagi.com/claims/huang-lex-2026","name":"Jensen Huang","role":"Co-founder and CEO","company":"NVIDIA","date":"2026-03-23","datePrecision":"day","quote":"I think it's now. I think we've achieved AGI.","paraphrase":false,"kind":"qualified","topic":"agi","models":["Claude"],"context":"The question defined AGI as AI able to build a billion-dollar company. Huang entertained a short-lived viral app reaching that scale, while distinguishing it from running a durable enterprise like NVIDIA. Claude was his illustrative example; he named no version.","source":"https://lexfridman.com/jensen-huang-transcript","sourceLabel":"Lex Fridman Podcast #494","evidence":"Primary transcript at 1:55:06–1:58:28; declaration at 1:56:31 and unversioned Claude example at 1:56:37. Episode published March 23.","expectedArrival":{"sourceDate":"2026-03-23","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"A present-tense claim under a special definition; no future forecast window."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2031211210205704461","url":"https://whocriedagi.com/claims/social-followup-x-2031211210205704461","name":"Aidan Clark","role":"AI researcher","company":"OpenAI","date":"2026-03-10","datePrecision":"day","quote":"When the book is written, AGI Day will be in today’s past.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"Clark offers a retrospective threshold claim without naming a system, providing an AGI definition or identifying the date that threshold was crossed.","source":"https://x.com/_aidan_clark_/status/2031211210205704461","sourceLabel":"Aidan Clark on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2031424856429117543","url":"https://whocriedagi.com/claims/social-followup-x-2031424856429117543","name":"Dean W. Ball","role":"AI policy researcher","company":"Independent","date":"2026-03-10","datePrecision":"day","quote":"I think we crossed the AGI line with the GPT 5.2/Opus 4.5/Gemini 3 generation of models","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-5.2","Claude Opus 4.5","Gemini 3"],"context":"Ball places his claimed threshold crossing in late November or early December 2025 and expects future historians to agree. The timeline records when he made the statement, in March 2026, rather than moving the quotation to the retrospective date.","source":"https://x.com/deanwball/status/2031424856429117543","sourceLabel":"Dean W. Ball on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2029123591871308272","url":"https://whocriedagi.com/claims/social-followup-x-2029123591871308272","name":"Elon Musk","role":"CEO","company":"Tesla","date":"2026-03-04","datePrecision":"day","quote":"Tesla will be one of the companies to make AGI and probably the first to make it in humanoid/atom-shaping form","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Musk predicts Tesla will build AGI, with a possible lead in humanoid or physical-world form. The post supplies no calendar deadline and names no particular robot or AI model.","source":"https://x.com/elonmusk/status/2029123591871308272","sourceLabel":"Elon Musk on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2026-03-04","kind":"unspecified","interpretation":"A predicted achievement by Tesla without an arrival date."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2025274779444654085","url":"https://whocriedagi.com/claims/tweet-sweep-x-2025274779444654085","name":"vals","role":"Commentator · @ValsTutor","company":"","date":"2026-02-21","datePrecision":"day","quote":"Whether technique novelty is needed for AGI is unknown, but I'd guess ~no, thus human level AGI in 3 years likely","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The @ValsTutor account calls human-level AGI in three years likely, based on a tentative belief that fundamentally novel techniques are unnecessary. The uncertainty about that premise remains part of the forecast.","source":"https://x.com/ValsTutor/status/2025274779444654085","sourceLabel":"vals on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2026-02-21","kind":"around","targetYear":2029,"approximate":true,"interpretation":"Human-level AGI roughly three years after February 2026, assuming no major technique novelty is required."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2024946760163242147","url":"https://whocriedagi.com/claims/tweet-sweep-x-2024946760163242147","name":"Carl Feynman","role":"Commentator · @carl_feynman","company":"","date":"2026-02-20","datePrecision":"day","quote":"Its raw intelligence is at human level.","paraphrase":false,"kind":"qualified","topic":"agi","models":["Claude Opus 4.6"],"context":"Feynman opens with “AGI is here” after using Claude Opus 4.6, describing it as a coworker with human-level raw intelligence. He immediately qualifies that assessment by saying it lacks wisdom and is easily misled. The official embed truncates the remaining caveat; no unseen wording is inferred.","source":"https://x.com/carl_feynman/status/2024946760163242147","sourceLabel":"Carl Feynman on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"cotra-2026-80000-explosion","url":"https://whocriedagi.com/claims/cotra-2026-80000-explosion","name":"Ajeya Cotra","role":"Senior advisor","company":"Open Philanthropy (now Coefficient Giving)","date":"2026-02-17","datePrecision":"day","quote":"After AI automates AI research, the remaining window before uncontrollably powerful superintelligence could be around a year or less.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"This is a conditional interval, not a prediction of an absolute arrival year. Recorded October 20, 2025, before her move to METR; date is public release.","source":"https://80000hours.org/podcast/episodes/ajeya-cotra-transformative-ai-crunch-time/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: White-knuckling the 12-month window after automated AI R&D, 00:47:21.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2026-dwarkesh","url":"https://whocriedagi.com/claims/amodei-2026-dwarkesh","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2026-02-13","datePrecision":"day","quote":"A country of AI geniuses has roughly even odds within one to three years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He discussed uncertainty around the arrival of powerful AI and distinguished technical capability from its subsequent economic deployment.","source":"https://www.dwarkesh.com/p/dario-amodei-2","sourceLabel":"Dwarkesh Podcast","evidence":"Primary interview published February 13; probabilistic timeline discussion.","target":"2027–2029, roughly 50% chance","expectedArrival":{"sourceDate":"2026-02-13","sourceDatePrecision":"day","kind":"range","startYear":2027,"endYear":2029,"classification":"ahead","interpretation":"The primary transcript gives a one-to-two, possibly one-to-three-year hunch, roughly even odds. This is a possible arrival range, not certainty or a claim of zero probability outside it.","probability":{"value":0.5,"scope":"forecast-horizon","approximate":true},"approximate":true,"evidence":"Primary transcript rechecked: discussion near 44 minutes distinguishes the short hunch from a much longer high-confidence horizon.","source":"https://www.dwarkesh.com/p/dario-amodei-2"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-2022086661170254203","url":"https://whocriedagi.com/claims/social-followup-x-2022086661170254203","name":"François Chollet","role":"Co-founder","company":"Ndea / ARC Prize Foundation","date":"2026-02-12","datePrecision":"day","quote":"AGI ~2030.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Chollet describes planned successive ARC benchmarks and says testing should continue until people can no longer propose tasks humans can do that AI cannot. The date is approximate, not a promise of a specific deadline.","source":"https://x.com/fchollet/status/2022086661170254203","sourceLabel":"François Chollet on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"kind":"around","sourceDate":"2026-02-12","targetYear":2030,"approximate":true,"interpretation":"Chollet writes AGI ~2030; the tilde denotes approximation."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2021011941578178714","url":"https://whocriedagi.com/claims/tweet-sweep-x-2021011941578178714","name":"Janus","role":"AI researcher · @repligate","company":"","date":"2026-02-10","datePrecision":"day","quote":"OpenAI built AGI first. In like 2020.","paraphrase":false,"kind":"arrived","topic":"agi","models":[],"context":"The pseudonymous researcher retrospectively places OpenAI’s achievement of AGI around 2020 and distinguishes it from a future system capable of runaway self-improvement. The event is dated to this February 2026 public statement, not backdated to 2020. The visible source does not name a model.","source":"https://x.com/repligate/status/2021011941578178714","sourceLabel":"Janus on X","evidence":"Discovered in Community Archive’s current public export and independently checked against X’s official public oEmbed. Author, post ID, visible excerpt and UTC publication date are verified. The statement records the speaker’s view, not this archive’s assessment of AGI.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2026-forbes-qualification","url":"https://whocriedagi.com/claims/altman-2026-forbes-qualification","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2026-02-03","datePrecision":"day","quote":"We basically have built AGI, or very close to it.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"Days after making the remark, he clarified that he meant it spiritually rather than literally and said additional breakthroughs were still required.","source":"https://www.forbes.com/sites/richardnieva/2026/02/03/sam-altman-explains-the-future/","sourceLabel":"Forbes","evidence":"Publication date is February 3; interview day is unspecified. Forbes Luxembourg's March 2 authorized republication preserves both the remark and its subsequent qualification.","expectedArrival":{"sourceDate":"2026-02-03","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"An achievement claim later described as nonliteral; no dated future horizon."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2026-adolescence","url":"https://whocriedagi.com/claims/amodei-2026-adolescence","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2026-01-26","datePrecision":"day","quote":"powerful AI could be as little as 1–2 years away","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He described systems exceeding human experts across many cognitive domains. He explicitly allowed a considerably longer timeline and framed rapid arrival as a serious possibility.","source":"https://darioamodei.com/essay/the-adolescence-of-technology","sourceLabel":"The Adolescence of Technology","evidence":"Author's January 2026 essay, released January 26.","target":"2027–2028 possible","expectedArrival":{"sourceDate":"2026-01-26","sourceDatePrecision":"day","kind":"range","startYear":2027,"endYear":2028,"classification":"ahead","interpretation":"One to two years describes a possible early-arrival scenario; the essay explicitly allows substantially later arrival. The upper year is not a hard deadline.","approximate":true,"confidence":"Possible early scenario; no numerical probability","upperBoundIsDeadline":false},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2026-davos","url":"https://whocriedagi.com/claims/musk-2026-davos","name":"Elon Musk","role":"Founder","company":"xAI","date":"2026-01-22","datePrecision":"day","quote":"We might have AI that is smarter than any human by the end of this year","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In conversation with Larry Fink, Musk made a possible-year-end forecast for exceeding individual human intelligence.","source":"https://www.weforum.org/stories/artificial-intelligence/elon-musk-technology-abundant-future-davos-2026/","sourceLabel":"World Economic Forum","evidence":"WEF's account of the January 22 Davos discussion.","target":"End of 2026 possible","expectedArrival":{"sourceDate":"2026-01-22","sourceDatePrecision":"day","kind":"by","deadlineYear":2026,"classification":"window-open","interpretation":"Possible arrival by the end of 2026; the deadline remains ahead of October 7.","deadlineDate":"2026-12-31","confidence":"Possible, not guaranteed"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-x-2014345509675155639","url":"https://whocriedagi.com/claims/social-x-2014345509675155639","name":"Shane Legg","role":"Cofounder and chief AGI scientist","company":"Google DeepMind","date":"2026-01-22","datePrecision":"day","quote":"AGI is approaching and will profoundly reshape the economy.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Legg advertises a senior economist role to study post-AGI economics. The phrase about AGI approaching has no numeric timetable and must not be converted to a specific arrival year.","source":"https://x.com/ShaneLegg/status/2014345509675155639","sourceLabel":"Shane Legg on X","evidence":"Author and statement text verified in X’s official public oEmbed response; UTC date computed from the original Snowflake post ID and checked against the embed’s date.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2026-davos","url":"https://whocriedagi.com/claims/hassabis-2026-davos","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2026-01-20","datePrecision":"day","quote":"There is a fifty percent chance of AGI by 2030.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At Bloomberg House in Davos, he emphasized a demanding definition, including scientific creativity and continued learning.","source":"https://www.investing.com/news/stock-market-news/google-deepmind-ceo-discusses-ai-progress-and-timeline-for-agi-93CH-4455818","sourceLabel":"Investing.com / Bloomberg House Davos","evidence":"January 20 reporting of the interview.","target":"2030, approximately 50% chance","expectedArrival":{"sourceDate":"2026-01-20","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"A cumulative 50% forecast by 2030, not a closed 2030 arrival window.","probability":{"value":0.5,"scope":"by-deadline","approximate":true}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2006774538990715280","url":"https://whocriedagi.com/claims/tweet-sweep-x-2006774538990715280","name":"Lisan al Gaib","role":"AI commentator · @scaling01","company":"","date":"2026-01-01","datePrecision":"day","quote":"My predictions for 2026: Coding and Mathematics AGI","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The pseudonymous account forecasts AGI-level abilities in coding and mathematics during 2026. The scope is domain-specific; it is not an unrestricted all-domain AGI prediction. The official embed truncates later parts of the post, so no additional forecast is inferred from them. A follow-up in the same thread puts coding superintelligence before general AGI, without dating general AGI.","source":"https://x.com/scaling01/status/2006774538990715280","sourceLabel":"Lisan al Gaib on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2026-01-01","kind":"by","deadlineYear":2026,"interpretation":"Coding and mathematics AGI during 2026.","note":"Domain-specific use of AGI; no claim of general competence across all domains."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-2006406873835253829","url":"https://whocriedagi.com/claims/tweet-sweep-x-2006406873835253829","name":"Fleek Futurist","role":"Commentator · @FleekFuturist","company":"","date":"2025-12-31","datePrecision":"day","quote":"Fleek Futurist reiterates a personal forecast of AGI by 2031, saying this has been their prediction for more than a decade.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The account contrasts its own 2031 forecast with a reported AI Futures Project model update. This record captures the account’s personal reiteration on December 31, 2025; it does not independently verify the other model or backdate the claim to an earlier publication.","source":"https://x.com/FleekFuturist/status/2006406873835253829","sourceLabel":"Fleek Futurist on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2025-12-31","kind":"by","deadlineYear":2031,"interpretation":"AGI by 2031, reiterated in December 2025."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2025-big-technology","url":"https://whocriedagi.com/claims/altman-2025-big-technology","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2025-12-18","datePrecision":"day","quote":"AGI may already have passed as a recognizable milestone.","paraphrase":true,"kind":"qualified","topic":"agi","models":[],"context":"Altman floated treating AGI as crossed and focusing on superintelligence, while acknowledging disagreement over what AGI means. He did not establish a shared test or name a qualifying model.","source":"https://podscripts.co/podcasts/big-technology-podcast/sam-altman-how-openai-wins-ai-buildout-logic-ipo-in-2026","sourceLabel":"Big Technology Podcast","evidence":"Transcript of Alex Kantrowitz's December 18 interview.","expectedArrival":{"sourceDate":"2025-12-18","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Speculation that the milestone may already have passed; no future arrival interval."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2025-ten-years","url":"https://whocriedagi.com/claims/altman-2025-ten-years","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2025-12-11","datePrecision":"day","quote":"In ten more years, I believe we are almost certain to build superintelligence.","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"The anniversary essay predicted a threshold beyond ordinary human-level AGI. It described a personal expectation rather than a demonstrated technical result.","source":"https://openai.com/index/ten-years/","sourceLabel":"OpenAI: Ten years","evidence":"OpenAI's December 11 anniversary post, authored by Altman.","target":"2035","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2025-deepmind-podcast","url":"https://whocriedagi.com/claims/legg-2025-deepmind-podcast","name":"Shane Legg","role":"Co-founder and Chief AGI Scientist","company":"Google DeepMind","date":"2025-12-11","datePrecision":"day","quote":"There remains a fifty percent chance of minimal AGI by 2028.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Legg distinguished a minimum general-intelligence threshold from the fuller capabilities that could follow later. He reaffirmed a forecast he had made many years earlier.","source":"https://www.youtube.com/watch?v=l3u_FAv33G0","sourceLabel":"Google DeepMind: The Podcast","evidence":"Official December 11 interview with Hannah Fry; transcript and Legg's own LinkedIn post corroborate the estimate.","target":"2028, approximately 50% chance","expectedArrival":{"sourceDate":"2025-12-11","sourceDatePrecision":"day","kind":"by","deadlineYear":2028,"classification":"window-open","interpretation":"A cumulative probability of minimal AGI by 2028. This does not predict arrival specifically during 2028.","probability":{"value":0.5,"scope":"by-deadline","approximate":true}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"sutskever-2025-dwarkesh","url":"https://whocriedagi.com/claims/sutskever-2025-dwarkesh","name":"Ilya Sutskever","role":"Co-founder and CEO","company":"Safe Superintelligence","date":"2025-11-25","datePrecision":"day","quote":"A human-like learner that becomes superhuman could arrive in five to twenty years.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"He emphasized research into generalization and learning, rather than treating larger versions of current systems as a sufficient solution.","source":"https://www.dwarkesh.com/p/ilya-sutskever-2","sourceLabel":"Dwarkesh Podcast","evidence":"Primary transcript, timeline exchange near 1:22:26; published November 25.","target":"2030–2045","expectedArrival":{"sourceDate":"2025-11-25","sourceDatePrecision":"day","kind":"range","startYear":2030,"endYear":2045,"classification":"ahead","interpretation":"Five to twenty years from the 2025 interview; a broad, approximate estimate for a human-like learner becoming superhuman.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1992254167394755023","url":"https://whocriedagi.com/claims/social-followup-x-1992254167394755023","name":"Eli Lifland","role":"Co-founder and research lead","company":"AI Futures Project","date":"2025-11-22","datePrecision":"day","quote":"My median AGI arrival estimate is now 2035.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Lifland publishes a table of his and Daniel Kokotajlo’s changing AGI medians, listing his own late-2025 value as 2035 after an early-2025 value of 2031. The primary table is compressed data rather than a sentence, so the display uses an explicit editorial paraphrase. Their coauthored January 2026 retrospective confirms that these are AGI median forecasts. Earlier rows are not added as separate historical posts.","source":"https://x.com/eli_lifland/status/1992254167394755023","sourceLabel":"Eli Lifland on X","evidence":"The original table and author were verified in X’s official public oEmbed. The coauthored January 2026 retrospective links this exact post as the source for Lifland’s November 2035 AGI median. UTC date was decoded from the post ID and checked against the embed date. Display text is an editorial paraphrase of the table.","expectedArrival":{"sourceDate":"2025-11-22","kind":"by","deadlineYear":2035,"approximate":true,"probability":{"value":0.5,"approximate":true,"scope":"AGI median arrival year"},"interpretation":"A median year of 2035, meaning an approximately 50% cumulative probability date, not certainty or a bounded interval."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1991564542103662729","url":"https://whocriedagi.com/claims/social-followup-x-1991564542103662729","name":"Daniel Kokotajlo","role":"Executive director","company":"AI Futures Project","date":"2025-11-20","datePrecision":"day","quote":"around 2030, lots of uncertainty though","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Kokotajlo says progress seems somewhat slower than the AI 2027 scenario and that his timeline was already later than 2027 when it was published. He describes his current public answer as around 2030 while explicitly retaining substantial uncertainty. His January 2026 retrospective identifies this as his November AGI median.","source":"https://x.com/DKokotajlo/status/1991564542103662729","sourceLabel":"Daniel Kokotajlo on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2025-11-20","kind":"around","targetYear":2030,"approximate":true,"interpretation":"Around 2030 with substantial uncertainty; do not treat this as a guaranteed deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"ng-2025-decades-or-longer","url":"https://whocriedagi.com/claims/ng-2025-decades-or-longer","name":"Andrew Ng","role":"Founder","company":"DeepLearning.AI","date":"2025-11-12","datePrecision":"day","quote":"AI able to do any human intellectual task was still decades away or longer, despite rapid practical progress.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Responding to a student's fear that AGI would eliminate meaningful future work, he distinguished today's general-purpose AI from universal intellectual competence.","source":"https://www.deeplearning.ai/the-batch/issue-327","sourceLabel":"The Batch — Andrew Ng's letter","evidence":"Author-signed opening letter in issue 327, published November 12, 2025.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-x-1981033485214089640","url":"https://whocriedagi.com/claims/social-x-1981033485214089640","name":"Yoshua Bengio","role":"Founder and scientific leader","company":"Mila / LawZero","date":"2025-10-22","datePrecision":"day","quote":"Frontier AI could match or exceed human capabilities within a few years, requiring safeguards and public participation.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"Bengio supports the Statement on Superintelligence while framing the potential timeline as a possibility. A few years is intentionally left unconverted to an exact year.","source":"https://x.com/Yoshua_Bengio/status/1981033485214089640","sourceLabel":"Yoshua Bengio on X","evidence":"Author and statement text verified in X’s official public oEmbed response; UTC date computed from the original Snowflake post ID and checked against the embed’s date.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2025-grok5-ten-percent","url":"https://whocriedagi.com/claims/musk-2025-grok5-ten-percent","name":"Elon Musk","role":"Founder","company":"xAI","date":"2025-10-18","datePrecision":"day","quote":"My estimate of the probability of Grok 5 achieving AGI is now at 10% and rising","paraphrase":false,"kind":"forecast","topic":"agi","models":["Grok 5"],"context":"This was an explicitly probabilistic assessment of an upcoming model.","source":"https://x.com/elonmusk/status/1979431839824777673","sourceLabel":"Elon Musk on X","evidence":"Exact wording, author and October 18, 2025 publication date verified in the original post’s official public X embed. The probability concerns Grok 5 achieving AGI; no calendar deadline is stated.","target":"Grok 5, estimated 10% chance","expectedArrival":{"sourceDate":"2025-10-18","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"The probability is attached to Grok 5 achieving AGI, not to a calendar deadline. A release date must not be substituted.","probability":{"value":0.1,"scope":"named-model-achieves-AGI","approximate":true},"confidence":"Estimate stated as rising"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"karpathy-2025-dwarkesh","url":"https://whocriedagi.com/claims/karpathy-2025-dwarkesh","name":"Andrej Karpathy","role":"Founder","company":"Eureka Labs","date":"2025-10-17","datePrecision":"day","quote":"It will take about a decade to work through all of those issues.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He was discussing the gap between current agents and dependable general digital workers, including learning and computer interaction. He presented the timeline as an intuition.","source":"https://www.dwarkesh.com/p/andrej-karpathy","sourceLabel":"Dwarkesh Podcast","evidence":"Primary interview published October 17.","target":"Around 2035","expectedArrival":{"sourceDate":"2025-10-17","sourceDatePrecision":"day","kind":"around","targetYear":2035,"classification":"unspecified","interpretation":"About a decade from 2025 gives a nominal 2035 marker. No tolerance band was specified, so exact window overlap is unknown.","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1978828394310619330","url":"https://whocriedagi.com/claims/tweet-sweep-x-1978828394310619330","name":"Dan Hendrycks","role":"AI researcher","company":"","date":"2025-10-16","datePrecision":"day","quote":"It seems like AGI won't arrive in a year, but it could easily this decade.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Hendrycks describes the remaining barriers to AGI as tractable. He doubts arrival within a year of October 2025 but regards arrival during the 2020s as plausible. This is not a probability estimate or a definite deadline.","source":"https://x.com/hendrycks/status/1978828394310619330","sourceLabel":"Dan Hendrycks on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2025-10-16","kind":"by","deadlineYear":2029,"approximate":true,"interpretation":"Could arrive in the 2020s; unlikely within one year of October 2025.","note":"Preserve both the near-term caveat and the uncertain decade-level forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1969732679269810502","url":"https://whocriedagi.com/claims/tweet-sweep-x-1969732679269810502","name":"kat","role":"Commentator · @transkatgirl","company":"","date":"2025-09-21","datePrecision":"day","quote":"Kat gives a 60% chance of AGI within five to twenty years and 90% within three to sixty years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The @transkatgirl account states two nested probability intervals in September 2025. The displayed interval is the 60% estimate, approximately 2030–2045; the wider 90% interval is approximately 2028–2085. The same post predicts widespread labor replacement within ten years after AGI, which is a separate conditional event.","source":"https://x.com/transkatgirl/status/1969732679269810502","sourceLabel":"kat on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","expectedArrival":{"sourceDate":"2025-09-21","kind":"range","startYear":2030,"endYear":2045,"approximate":true,"probability":{"value":0.6},"interpretation":"60% probability within five to twenty years after September 2025.","note":"The post also gives a wider 90% interval of three to sixty years. Neither interval is a guaranteed deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2025-grok5-chance","url":"https://whocriedagi.com/claims/musk-2025-grok5-chance","name":"Elon Musk","role":"Founder","company":"xAI","date":"2025-09-17","datePrecision":"day","quote":"I now think @xAI has a chance of reaching AGI with @Grok 5.","paraphrase":false,"kind":"forecast","topic":"agi","models":["Grok 5"],"context":"Musk described a possibility for an upcoming model rather than an accomplished result.","source":"https://x.com/elonmusk/status/1968202372276163029","sourceLabel":"Elon Musk on X","evidence":"September 17 post, reproduced in contemporaneous reporting.","target":"Grok 5","expectedArrival":{"sourceDate":"2025-09-17","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Grok 5 is a model milestone, not a dated arrival window."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2025-athens","url":"https://whocriedagi.com/claims/hassabis-2025-athens","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2025-09-12","datePrecision":"day","quote":"AGI could arrive within the next decade.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Speaking in Athens, he connected the forecast to scientific progress and the need to prepare people for changes in work.","source":"https://apnews.com/article/85bff114c30cbea4b951ab93dcc1e6d1","sourceLabel":"Associated Press / Athens","evidence":"AP report of his September 12 appearance.","target":"By approximately 2035","expectedArrival":{"sourceDate":"2025-09-12","sourceDatePrecision":"day","kind":"by","deadlineYear":2035,"classification":"window-open","interpretation":"Within the next decade supplies an approximate upper horizon from 2025, not a ten-year-wide predicted arrival range.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1958499441469739329","url":"https://whocriedagi.com/claims/social-followup-x-1958499441469739329","name":"Elon Musk","role":"Founder","company":"xAI","date":"2025-08-21","datePrecision":"day","quote":"Wait until you see Grok 5. I think it has a shot at being true AGI.","paraphrase":false,"kind":"forecast","topic":"agi","models":["Grok 5"],"context":"Musk identifies Grok 5 as a possible AGI and says he has not felt that about anything before. The chance is not quantified and no launch date is given. This August statement precedes his separately archived September and October claims.","source":"https://x.com/elonmusk/status/1958499441469739329","sourceLabel":"Elon Musk on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2025-08-21","kind":"unspecified","interpretation":"A possible AGI outcome for Grok 5, without a calendar target or numerical probability."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1958229951536316528","url":"https://whocriedagi.com/claims/social-followup-x-1958229951536316528","name":"Daniel Kokotajlo","role":"Executive director","company":"AI Futures Project","date":"2025-08-20","datePrecision":"day","quote":"my bottom-line numbers are somewhat different (median ~EOY 2029)","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In an explicit discussion of AGI timelines with Ryan Greenblatt, Kokotajlo places his own median near the end of 2029. A median is a 50% cumulative probability date, not a claim that arrival cannot occur earlier or later. This is a dated update after AI 2027’s publication.","source":"https://x.com/DKokotajlo/status/1958229951536316528","sourceLabel":"Daniel Kokotajlo on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2025-08-20","kind":"by","deadlineYear":2029,"approximate":true,"probability":{"value":0.5,"approximate":true,"scope":"AGI by approximately the end of 2029"},"interpretation":"Approximate median at the end of 2029, not a bounded arrival interval."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1955790948371456085","url":"https://whocriedagi.com/claims/tweet-sweep-x-1955790948371456085","name":"Janus","role":"Commentator · @repligate","company":"","date":"2025-08-14","datePrecision":"day","quote":"gpt-4o is AGI, its just not the smartest one","paraphrase":false,"kind":"qualified","topic":"agi","models":["GPT-4o"],"context":"The @repligate account calls GPT-4o AGI while qualifying its relative intelligence. This is a model-specific claim distinct from the account’s later retrospective claim about OpenAI in 2020.","source":"https://x.com/repligate/status/1955790948371456085","sourceLabel":"Janus on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hinton-2025-ai4","url":"https://whocriedagi.com/claims/hinton-2025-ai4","name":"Geoffrey Hinton","role":"Professor Emeritus","company":"University of Toronto","date":"2025-08-12","datePrecision":"day","quote":"AGI could arrive in five to twenty years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At Ai4 in Las Vegas, he described this as a reasonable range and said further breakthroughs could still be needed.","source":"https://www.reviewjournal.com/?p=3427271","sourceLabel":"Las Vegas Review-Journal / Ai4","evidence":"Reporting on his August 12 keynote.","target":"2030–2045","expectedArrival":{"sourceDate":"2025-08-12","sourceDatePrecision":"day","kind":"range","startYear":2030,"endYear":2045,"classification":"ahead","interpretation":"Five to twenty years from 2025; an approximate estimate that still depended on further breakthroughs.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"zuckerberg-2025-superintelligence","url":"https://whocriedagi.com/claims/zuckerberg-2025-superintelligence","name":"Mark Zuckerberg","role":"Founder and CEO","company":"Meta","date":"2025-07-30","datePrecision":"day","quote":"Developing superintelligence is now in sight.","paraphrase":false,"kind":"qualified","topic":"asi","models":[],"context":"He pointed to early signs of AI systems improving themselves, while describing that progress as slow. His letter did not announce a completed AGI system or a specific deadline.","source":"https://www.meta.com/superintelligence/","sourceLabel":"Personal Superintelligence","evidence":"Zuckerberg's July 30 public letter.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2025-lex","url":"https://whocriedagi.com/claims/hassabis-2025-lex","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2025-07-23","datePrecision":"day","quote":"There is a fifty percent chance of AGI by 2030.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His threshold included inventing new scientific explanations and other broadly general abilities. He expected additional breakthroughs to be necessary.","source":"https://lexfridman.com/demis-hassabis-2-transcript/","sourceLabel":"Lex Fridman Podcast #475","evidence":"Primary transcript, timeline discussion at 00:52:33.","target":"2030, approximately 50% chance","expectedArrival":{"sourceDate":"2025-07-23","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"Cumulative 50% probability by 2030; no lower bound on arrival was specified.","probability":{"value":0.5,"scope":"by-deadline","approximate":true}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1947087465090851063","url":"https://whocriedagi.com/claims/social-followup-x-1947087465090851063","name":"Elon Musk","role":"Founder","company":"xAI","date":"2025-07-21","datePrecision":"day","quote":"The continuous RL improvement of Grok feels like AGI.","paraphrase":false,"kind":"qualified","topic":"agi","models":["Grok 4"],"context":"Musk says Grok’s continuing reinforcement-learning improvement feels like AGI and adds that Grok 4 is smarter than it was a few days earlier. The experiential “feels like” wording is preserved as a qualified claim, not a categorical declaration.","source":"https://x.com/elonmusk/status/1947087465090851063","sourceLabel":"Elon Musk on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1946558716695748764","url":"https://whocriedagi.com/claims/social-followup-x-1946558716695748764","name":"Emad Mostaque","role":"Founder","company":"Intelligent Internet","date":"2025-07-19","datePrecision":"day","quote":"AGI is already here. All the components exist; we just need to stitch them together.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"Mostaque explicitly says AGI is already here, but immediately qualifies this by saying that existing components still need to be stitched together. Kept as a qualified arrival claim rather than evidence of an assembled, generally capable system. No model is named.","source":"https://x.com/EMostaque/status/1946558716695748764","sourceLabel":"Emad Mostaque on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1944658970553307435","url":"https://whocriedagi.com/claims/social-followup-x-1944658970553307435","name":"Elon Musk","role":"Founder","company":"xAI","date":"2025-07-14","datePrecision":"day","quote":"Probably smarter than any single human at anything in <2 years","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Musk distinguishes current AI from a future system surpassing any individual human across tasks. The verified excerpt predicts that threshold in under two years from July 2025. The official embed truncates a later claim about all humans combined; that later target is deliberately excluded from this record.","source":"https://x.com/elonmusk/status/1944658970553307435","sourceLabel":"Elon Musk on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","originalTerminology":"smarter than any single human at anything","expectedArrival":{"sourceDate":"2025-07-14","kind":"by","deadlineYear":2027,"approximate":true,"interpretation":"Less than two years from July 14, 2025, approximately before mid-July 2027.","note":"The date concerns surpassing any single human across tasks; the truncated collective-human comparison is not used."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"greenblatt-2025-80000-automation","url":"https://whocriedagi.com/claims/greenblatt-2025-80000-automation","name":"Ryan Greenblatt","role":"Chief scientist","company":"Redwood Research","date":"2025-07-08","datePrecision":"day","quote":"There was a 25% chance of AI fully automating an AI company within four years, and a 50% chance within eight years.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"The forecast concerns AI R&D and company operations, rather than every physical task. He separated capability availability from deployment. Recorded February 21, 2025; intervals run from that conversation.","source":"https://80000hours.org/podcast/episodes/ryan-greenblatt-ai-automation-sabotage-takeover/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: How close are we to automating AI R&D?, 00:01:29.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2025-gentle-singularity","url":"https://whocriedagi.com/claims/altman-2025-gentle-singularity","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2025-06-10","datePrecision":"day","quote":"Humanity is close to building digital superintelligence","paraphrase":false,"kind":"qualified","topic":"asi","models":[],"context":"The essay described current systems as approaching AGI and helping accelerate AI research. It also acknowledged their remaining limitations and did not identify one model as conclusively AGI.","source":"https://blog.samaltman.com/the-gentle-singularity","sourceLabel":"The Gentle Singularity","evidence":"Altman's June 10 essay.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1930633754801737798","url":"https://whocriedagi.com/claims/tweet-sweep-x-1930633754801737798","name":"redJ","role":"Commentator · @sudoredj","company":"","date":"2025-06-05","datePrecision":"day","quote":"we have AGI. our best models pass the turing test, this is AGI.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"The @sudoredj account argues that passing the Turing test is sufficient to call current models AGI and asks for different terminology for stronger requirements. The arrival claim depends on that definition; no particular model is named.","source":"https://x.com/sudoredj/status/1930633754801737798","sourceLabel":"redJ on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-wired-2025","url":"https://whocriedagi.com/claims/hassabis-wired-2025","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2025-06-03","datePrecision":"day","quote":"There is roughly a fifty percent chance of AGI in five to ten years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"This was a probabilistic estimate at WIRED's Big Interview event, not a guaranteed arrival date.","source":"https://www.wired.com/video/watch/big-interview-demis-hassabis-on-the-future-of-work-in-the-age-of-ai","sourceLabel":"WIRED: The Big Interview","evidence":"WIRED's June 3 event video.","target":"2030–2035, approximately 50% chance","expectedArrival":{"sourceDate":"2025-06-03","sourceDatePrecision":"day","kind":"range","startYear":2030,"endYear":2035,"classification":"ahead","interpretation":"The stated five-to-ten-year forecast horizon maps to 2030–2035. The roughly 50% estimate expresses uncertainty, not a guarantee or exclusion of earlier arrival.","probability":{"value":0.5,"scope":"forecast-horizon","approximate":true},"approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"barnes-2025-80000-self-improvement","url":"https://whocriedagi.com/claims/barnes-2025-80000-self-improvement","name":"Beth Barnes","role":"CEO","company":"METR","date":"2025-06-02","datePrecision":"day","quote":"Autonomous AI self-improvement could be possible within two years; even substantially shorter timelines were hard to exclude.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"She discussed the risk of an intelligence explosion from automated AI research. This was a possibility claim, with a small probability assigned to much shorter timelines.","source":"https://80000hours.org/podcast/episodes/beth-barnes-ai-safety-evals/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Publisher introduction quotes her forecast; transcript section 01:07:55.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"brin-2025-google-io","url":"https://whocriedagi.com/claims/brin-2025-google-io","name":"Sergey Brin","role":"Co-founder","company":"Google","date":"2025-05-20","datePrecision":"day","quote":"AGI just before 2030.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Google I/O interview.","source":"https://www.axios.com/2025/05/21/google-sergey-brin-demis-hassabis-agi-2030","sourceLabel":"Axios","evidence":"Reported May 21.","target":"Before 2030","expectedArrival":{"sourceDate":"2025-05-20","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"Just before 2030 is a deadline-side statement with an unspecified distance before it; no start year is inferred.","deadlineRelation":"before","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-mollick-2025-jagged-agi","url":"https://whocriedagi.com/claims/social-followup-mollick-2025-jagged-agi","name":"Ethan Mollick","role":"Professor","company":"Wharton School, University of Pennsylvania","date":"2025-04-20","datePrecision":"day","quote":"I really don’t know, but I do think they can be credibly seen as a form of “Jagged AGI”","paraphrase":false,"kind":"qualified","topic":"agi","models":["o3","Gemini 2.5 Pro"],"context":"Asked within his essay whether o3 and Gemini 2.5 are AGI, Mollick explicitly says he does not know because of definitional problems, then proposes the qualified phrase “Jagged AGI.” He emphasizes uneven capabilities and continuing need for human expertise. The uncertainty remains in the excerpt.","source":"https://www.oneusefulthing.org/p/on-jagged-agi-o3-gemini-25-and-everything","sourceLabel":"Ethan Mollick · One Useful Thing","evidence":"Exact excerpt, author and publication timestamp verified through the publication’s own public post API. The original essay supplies both the AGI qualification and the named models.","originalTerminology":"Jagged AGI","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"besiroglu-2025-dwarkesh-remote-worker","url":"https://whocriedagi.com/claims/besiroglu-2025-dwarkesh-remote-worker","name":"Tamay Besiroglu","role":"Cofounder","company":"Mechanize","date":"2025-04-17","datePrecision":"day","quote":"His timeline for AI doing every remote job was about five years earlier than Ege Erdil's 2045 estimate.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He distinguished doing literally everything remotely from doing most remote tasks. The implied approximately 2040 date is an interpretation of his comparison.","source":"https://www.dwarkesh.com/p/ege-tamay","sourceLabel":"Dwarkesh Podcast — original interview transcript","evidence":"Opening timeline exchange, 00:02:11–00:02:22.","target":"Around 2040, implied by comparison","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"erdil-2025-dwarkesh-remote-worker","url":"https://whocriedagi.com/claims/erdil-2025-dwarkesh-remote-worker","name":"Ege Erdil","role":"Cofounder","company":"Mechanize","date":"2025-04-17","datePrecision":"day","quote":"A drop-in AI replacement able to do every kind of remote work seemed likely around 2045.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He used complete remote-work replacement as his threshold, substantially beyond being useful for many tasks.","source":"https://www.dwarkesh.com/p/ege-tamay","sourceLabel":"Dwarkesh Podcast — original interview transcript","evidence":"Opening timeline exchange, 00:02:05–00:02:21.","target":"2045","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-cowen-2025-o3-agi","url":"https://whocriedagi.com/claims/social-followup-cowen-2025-o3-agi","name":"Tyler Cowen","role":"Professor of economics","company":"George Mason University","date":"2025-04-16","datePrecision":"day","quote":"I think it is AGI, seriously.","paraphrase":false,"kind":"arrived","topic":"agi","models":["o3"],"context":"Cowen’s authored post is titled “o3 and AGI, is April 16th AGI day?” He explicitly calls the newly released model AGI while arguing that integration and social effects will take time. His claim is personal judgment rather than a formal benchmark definition.","source":"https://marginalrevolution.com/marginalrevolution/2025/04/o3-and-agi-is-april-16th-agi-day.html","sourceLabel":"Tyler Cowen · Marginal Revolution","evidence":"Exact excerpt, named byline and April 16 publication date verified on the author’s original post. The HTML publication metadata gives the matching UTC timestamp.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2025-time-interview","url":"https://whocriedagi.com/claims/hassabis-2025-time-interview","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2025-04-15","datePrecision":"day","quote":"AGI remains roughly five to ten years away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He linked AGI to genuinely general capabilities, including scientific creativity, rather than simply performing existing workplace tasks.","source":"https://time.com/7277608/demis-hassabis-interview-time100-2025/","sourceLabel":"TIME interview","evidence":"Published April 15, 2025.","target":"2030–2035","expectedArrival":{"sourceDate":"2025-04-15","sourceDatePrecision":"day","kind":"range","startYear":2030,"endYear":2035,"classification":"ahead","interpretation":"Roughly five to ten years from the 2025 interview.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1909940254368530446","url":"https://whocriedagi.com/claims/tweet-sweep-x-1909940254368530446","name":"John David Pressman","role":"Commentator · @jd_pressman","company":"","date":"2025-04-09","datePrecision":"day","quote":"I don't think it's unreasonable to speculate that we may have ASI in 2027","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"Pressman says ASI in 2027 is a reasonable possibility to speculate about. He does not present 2027 as his median, promise, or high-confidence expectation.","source":"https://x.com/jd_pressman/status/1909940254368530446","sourceLabel":"John David Pressman on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2025-04-09","kind":"around","targetYear":2027,"interpretation":"ASI in 2027 is presented as a possibility, without a probability.","note":"A speculative horizon, not a confident deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1908152814301311303","url":"https://whocriedagi.com/claims/tweet-sweep-x-1908152814301311303","name":"Benjamin Todd","role":"Writer","company":"","date":"2025-04-04","datePrecision":"day","quote":"Todd says AGI before 2030 is difficult to rule out, while explicitly emphasizing that its arrival is uncertain.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Todd argues that an extension of current trends could produce AGI before 2030. This expresses a possibility rather than a best estimate, median forecast or promise.","source":"https://x.com/ben_j_todd/status/1908152814301311303","sourceLabel":"Benjamin Todd on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2025-04-04","kind":"by","deadlineYear":2030,"deadlineRelation":"before","interpretation":"AGI before 2030 is possible; the speaker does not give a probability.","note":"A possible horizon, not a confident deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kokotajlo-2025-ai-2027","url":"https://whocriedagi.com/claims/kokotajlo-2025-ai-2027","name":"Daniel Kokotajlo","role":"Co-founder and Director","company":"AI Futures Project","date":"2025-04-03","datePrecision":"day","quote":"2027 is a plausible scenario for AGI followed by rapid superintelligence.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"AI 2027 was a collaboratively authored scenario, not a promise or a 50% deadline. The authors distinguished a leading scenario from median timelines, which could be later.","source":"https://ai-2027.com/","sourceLabel":"AI 2027","evidence":"Scenario published April 3; authored with Scott Alexander, Thomas Larsen, Eli Lifland and Romeo Dean.","target":"2027 scenario","expectedArrival":{"sourceDate":"2025-04-03","sourceDatePrecision":"day","kind":"around","targetYear":2027,"classification":"unspecified","interpretation":"2027 is the year of a scenario, not a cumulative deadline, median or guaranteed outcome. No reliable interval width can be extracted.","confidence":"Scenario / leading possibility; no numerical probability","nominalYearRelation":"ahead","estimateType":"scenario"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"schmidt-2025-scsp","url":"https://whocriedagi.com/claims/schmidt-2025-scsp","name":"Eric Schmidt","role":"Co-founder; former Google CEO","company":"Special Competitive Studies Project","date":"2025-04-01","datePrecision":"month","quote":"AGI could match the strongest human specialists within three to five years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He described a system competitive with leading people in mathematics, physics, art, writing and other fields. He separately placed superintelligence around six years away.","source":"https://www.musicbusinessworldwide.com/ex-google-ceo-eric-schmidt-ai-that-is-as-smart-as-the-smartest-artist-will-be-here-in-3-to-5-years/","sourceLabel":"Music Business Worldwide / SCSP","evidence":"April 16 reporting of his discussion with Jeanne Meserve; exact event day not established.","target":"2028–2030","expectedArrival":{"sourceDate":"2025-04-01","sourceDatePrecision":"month","kind":"range","startYear":2028,"endYear":2030,"classification":"ahead","interpretation":"The three-to-five-year horizon from 2025 is retained as an approximate arrival range, not a claim that AGI is impossible earlier.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1902182509141479598","url":"https://whocriedagi.com/claims/tweet-sweep-x-1902182509141479598","name":"Christopher C. Smith","role":"Commentator · @christophcsmith","company":"","date":"2025-03-19","datePrecision":"day","quote":"Smith calls Claude 3.7 AGI for autonomous coding work, while emphasizing its tendency to fixate on details and lose sight of goals.","paraphrase":true,"kind":"qualified","topic":"agi","models":["Claude 3.7"],"context":"Smith’s definition is based on autonomous codebase changes and implementing features end to end at better-than-intern level. His stated limitations remain attached to the arrival claim; it is not an unrestricted claim of reliable performance in every domain.","source":"https://x.com/christophcsmith/status/1902182509141479598","sourceLabel":"Christopher C. Smith on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2025-cnbc","url":"https://whocriedagi.com/claims/hassabis-2025-cnbc","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2025-03-17","datePrecision":"day","quote":"Human-level AI could be here in five to ten years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The estimate concerned broadly human-level capabilities, not a claim that an existing chatbot was already AGI.","source":"https://www.cnbc.com/2025/03/17/human-level-ai-will-be-here-in-5-to-10-years-deepmind-ceo-says.html","sourceLabel":"CNBC","evidence":"CNBC interview published March 17.","target":"2030–2035","expectedArrival":{"sourceDate":"2025-03-17","sourceDatePrecision":"day","kind":"range","startYear":2030,"endYear":2035,"classification":"ahead","interpretation":"Five to ten years from 2025; year precision only.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1896248099968098461","url":"https://whocriedagi.com/claims/tweet-sweep-x-1896248099968098461","name":"vals","role":"Commentator · @ValsTutor","company":"","date":"2025-03-02","datePrecision":"day","quote":"I'm pretty sure we'll have cheaper than human AGI before 1 decade passes","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The @ValsTutor account expects AGI that is cheaper than human labor within a decade of March 2025 and substantial societal change. The cost qualification is part of the prediction.","source":"https://x.com/ValsTutor/status/1896248099968098461","sourceLabel":"vals on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2025-03-02","kind":"by","deadlineYear":2035,"approximate":true,"interpretation":"Cheaper-than-human AGI in less than a decade after March 2025."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1895830156780171758","url":"https://whocriedagi.com/claims/tweet-sweep-x-1895830156780171758","name":"xlr8harder","role":"Commentator · @xlr8harder","company":"","date":"2025-03-01","datePrecision":"day","quote":"agi is already here, but it's jagged and doesn't fit in well with our preconceptions.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"The account calls AGI present but uneven, while expecting at least several more years before systems can function independently in most human roles. The claim therefore separates general intelligence from reliable autonomy.","source":"https://x.com/xlr8harder/status/1895830156780171758","sourceLabel":"xlr8harder on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"brin-2025-final-race-memo","url":"https://whocriedagi.com/claims/brin-2025-final-race-memo","name":"Sergey Brin","role":"Co-founder","company":"Google","date":"2025-02-26","datePrecision":"day","quote":"the final race to A.G.I. is afoot","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In a memo to the Gemini team, Brin presented AGI as an immediate competitive race. He did not attach a firm date or declare a model had achieved it.","source":"https://9to5google.com/2025/02/27/sergey-brin-google-agi/","sourceLabel":"9to5Google / reported internal memo","evidence":"February 27 report quoting the memo, reported as sent Wednesday.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2025-three-observations","url":"https://whocriedagi.com/claims/altman-2025-three-observations","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2025-02-09","datePrecision":"day","quote":"Systems that start to point toward AGI are coming into view.","paraphrase":true,"kind":"qualified","topic":"agi","models":[],"context":"Altman described approaching a threshold, not having conclusively crossed it. His working definition involved handling complex problems at human level across many fields.","source":"https://blog.samaltman.com/three-observations","sourceLabel":"Three Observations","evidence":"Altman's February 9 essay.","expectedArrival":{"sourceDate":"2025-02-09","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Approaching a threshold supplies no date. Near-term agent examples are not an AGI arrival deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"son-2025-tokyo","url":"https://whocriedagi.com/claims/son-2025-tokyo","name":"Masayoshi Son","role":"Chairman and CEO","company":"SoftBank Group","date":"2025-02-03","datePrecision":"day","quote":"AGI will arrive much sooner than two years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"First deployed in large Japanese companies.","source":"https://www.axios.com/2025/02/03/altman-ai-agents-openai-agi-japan","sourceLabel":"Axios","evidence":"Tokyo event.","target":"Before 2027","expectedArrival":{"sourceDate":"2025-02-03","sourceDatePrecision":"day","kind":"by","deadlineYear":2027,"classification":"window-open","interpretation":"Much sooner than two years from February 2025 supplies an outer 2027 horizon; it does not establish an arrival interval.","deadlineRelation":"before-relative-anniversary","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2025-export-controls","url":"https://whocriedagi.com/claims/amodei-2025-export-controls","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2025-01-29","datePrecision":"day","quote":"AI smarter than almost all people at almost everything is most likely in 2026–2027.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He argued that DeepSeek's advances did not change this forecast. His projection depended on very large investments in chips and computing infrastructure.","source":"https://darioamodei.com/post/on-deepseek-and-export-controls","sourceLabel":"On DeepSeek and Export Controls","evidence":"Amodei's January 29 essay.","target":"2026–2027","expectedArrival":{"sourceDate":"2025-01-29","sourceDatePrecision":"day","kind":"range","startYear":2026,"endYear":2027,"classification":"window-open","interpretation":"The author explicitly identified 2026–2027 as the most likely period. October 2026 is within that named-year range.","confidence":"Most likely period; no numerical probability"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1884576969699770528","url":"https://whocriedagi.com/claims/tweet-sweep-x-1884576969699770528","name":"Tomas Pueyo","role":"Writer","company":"","date":"2025-01-29","datePrecision":"day","quote":"We're 1-6 years away from AGI","paraphrase":false,"kind":"forecast","topic":"agi","models":["OpenAI o3","DeepSeek"],"context":"Pueyo offers his own synthesis of prediction markets, insider statements, benchmarks and recent model progress. The one-to-six-year range is relative to January 2025. It is not an official forecast by a model developer.","source":"https://x.com/tomaspueyo/status/1884576969699770528","sourceLabel":"Tomas Pueyo on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2025-01-29","kind":"range","startYear":2026,"endYear":2031,"approximate":true,"interpretation":"One to six years after January 2025, expressed at calendar-year precision."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2025-davos-wsj","url":"https://whocriedagi.com/claims/amodei-2025-davos-wsj","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2025-01-21","datePrecision":"day","quote":"AI could surpass almost everyone at almost every task within two to three years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At Davos, he said the exact timing was uncertain, but he did not expect a long delay beyond 2027.","source":"https://arstechnica.com/ai/2025/01/anthropic-chief-says-ai-could-surpass-almost-all-humans-at-almost-everything-shortly-after-2027/","sourceLabel":"Ars Technica / WSJ Davos interview","evidence":"January 22 report of Tuesday's Journal House discussion.","target":"2027, or shortly afterward","expectedArrival":{"sourceDate":"2025-01-21","sourceDatePrecision":"day","kind":"around","targetYear":2027,"classification":"unspecified","interpretation":"2027 or shortly afterward gives a reference year with an undefined later tail. Do not invent 2028 or another firm ending.","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-reflections-2025","url":"https://whocriedagi.com/claims/altman-reflections-2025","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2025-01-05","datePrecision":"day","quote":"We are now confident we know how to build AGI as we have traditionally understood it.","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"He expressed confidence in a path to building AGI, rather than announcing that it had already been completed. The essay also anticipated agents joining the workforce in 2025.","source":"https://blog.samaltman.com/reflections","sourceLabel":"Reflections","evidence":"Altman's January 5 essay; some archive timestamps use January 6 UTC.","expectedArrival":{"sourceDate":"2025-01-05","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Confidence in knowing how to build AGI is not a forecast date. The separate 2025 workforce-agent discussion does not date AGI."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1874891358269592063","url":"https://whocriedagi.com/claims/tweet-sweep-x-1874891358269592063","name":"Tenobrus","role":"Commentator · @tenobrus","company":"","date":"2025-01-02","datePrecision":"day","quote":"we will have AGI in 4 years","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Tenobrus forecasts AGI four years after January 2025 in a comparison with the expected pace of fully autonomous robotaxis. The post supplies no AGI definition or probability.","source":"https://x.com/tenobrus/status/1874891358269592063","sourceLabel":"Tenobrus on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2025-01-02","kind":"around","targetYear":2029,"approximate":true,"interpretation":"Four years after January 2025."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1870476568663048624","url":"https://whocriedagi.com/claims/tweet-sweep-x-1870476568663048624","name":"Fleek Futurist","role":"Commentator · @FleekFuturist","company":"","date":"2024-12-21","datePrecision":"day","quote":"GPT-4 was AGI to me.","paraphrase":false,"kind":"qualified","topic":"agi","models":["GPT-4"],"context":"The account makes an explicitly personal assessment, citing GPT-4’s effect on their career, habits and productivity. “To me” is part of the claim, rather than a statement of consensus.","source":"https://x.com/FleekFuturist/status/1870476568663048624","sourceLabel":"Fleek Futurist on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"suleyman-2024-decoder","url":"https://whocriedagi.com/claims/suleyman-2024-decoder","name":"Mustafa Suleyman","role":"CEO, Microsoft AI","company":"Microsoft","date":"2024-12-09","datePrecision":"day","quote":"AGI is plausible within five to ten years, with substantial uncertainty.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He discussed several future hardware generations and rejected high confidence in a two-year timeline. He distinguished a general-purpose human-level learner from recursively improving superintelligence.","source":"https://www.theverge.com/24314821/microsoft-ai-ceo-mustafa-suleyman-google-deepmind-openai-inflection-agi-decoder-podcast","sourceLabel":"The Verge: Decoder","evidence":"Primary interview transcript published December 9.","target":"Approximately 2029–2034","expectedArrival":{"sourceDate":"2024-12-09","sourceDatePrecision":"day","kind":"range","startYear":2029,"endYear":2034,"classification":"ahead","interpretation":"The interview's uncertain, hardware-generation-based five-to-ten-year horizon maps approximately to these years.","approximate":true,"confidence":"Substantial uncertainty; not high confidence in a two-year arrival"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1865820720900739372","url":"https://whocriedagi.com/claims/tweet-sweep-x-1865820720900739372","name":"Oliver Habryka","role":"Commentator · @ohabryka","company":"","date":"2024-12-08","datePrecision":"day","quote":"It was talking about predicting AGI within 10-15 years (which is also roughly my prediction).","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Habryka endorses a rough ten-to-fifteen-year AGI horizon in December 2024, while allowing for progress to slow if the compute overhang is exhausted. “Within” supplies an upper horizon; it does not rule out earlier arrival.","source":"https://x.com/ohabryka/status/1865820720900739372","sourceLabel":"Oliver Habryka on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2024-12-08","kind":"by","deadlineYear":2039,"approximate":true,"interpretation":"Roughly within ten to fifteen years of December 2024.","note":"The fifteen-year upper horizon is shown; no earliest-arrival boundary is asserted."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hinton-2024-nobel","url":"https://whocriedagi.com/claims/hinton-2024-nobel","name":"Geoffrey Hinton","role":"Professor Emeritus","company":"University of Toronto","date":"2024-12-07","datePrecision":"day","quote":"Superintelligence may arrive in five to twenty years.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"At a Nobel press conference, Hinton described a much shorter horizon than he had previously expected. The estimate concerns AI exceeding people, not a claim about an existing model.","source":"https://toronto.citynews.ca/2024/12/08/nobel-recipient-geoffrey-hinton-wishes-he-thoughts-of-ai-safety-sooner/","sourceLabel":"Canadian Press / Nobel press conference","evidence":"December 8 reporting identifies the press conference as Saturday, December 7.","target":"2029–2044","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1858656214562766949","url":"https://whocriedagi.com/claims/tweet-sweep-x-1858656214562766949","name":"Nathan Young","role":"Forecaster","company":"","date":"2024-11-18","datePrecision":"day","quote":"I predict AGI in the next 10 years (65%)","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Young gives a 65% probability of AGI within ten years of November 2024. This is a probabilistic forecast, not a claim that arrival is guaranteed.","source":"https://x.com/NathanpmYoung/status/1858656214562766949","sourceLabel":"Nathan Young on X","evidence":"Discovered in Community Archive’s current public export and independently checked against X’s official public oEmbed. Author, post ID, visible excerpt and UTC publication date are verified. The statement records the speaker’s view, not this archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2024-11-18","kind":"by","deadlineYear":2034,"approximate":true,"probability":{"value":0.65},"interpretation":"65% chance within ten years of November 2024, shown at year precision."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-lex-2024","url":"https://whocriedagi.com/claims/amodei-lex-2024","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2024-11-11","datePrecision":"day","quote":"Extrapolating current progress suggests powerful AI in 2026 or 2027, possibly with some delay.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Amodei stressed that this was an uncertain extrapolation, not a scientific guarantee. Data, computing capacity and geopolitical disruption could derail it.","source":"https://lexfridman.com/dario-amodei-transcript/","sourceLabel":"Lex Fridman Podcast #452","evidence":"Primary transcript, AGI timeline discussion at 2:18:36–2:20:39.","target":"2026–2027, with possible delays","expectedArrival":{"sourceDate":"2024-11-11","sourceDatePrecision":"day","kind":"range","startYear":2026,"endYear":2027,"classification":"window-open","interpretation":"These are explicit calendar years from a trend extrapolation. He expected possible delays, so 2027 is not a hard deadline.","approximate":true,"upperBoundIsDeadline":false},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2024-yc-next-year","url":"https://whocriedagi.com/claims/altman-2024-yc-next-year","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2024-11-08","datePrecision":"day","quote":"AGI.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"This was his answer when asked what he was most excited about in 2025. It was a brief expression of expectation, not a detailed technical prediction.","source":"https://www.youtube.com/watch?v=xXCBz_8hM9w","sourceLabel":"Y Combinator: How to Build the Future","evidence":"Official interview's closing exchange; published November 8.","target":"2025","expectedArrival":{"sourceDate":"2024-11-08","sourceDatePrecision":"day","kind":"range","startYear":2025,"endYear":2025,"classification":"window-elapsed","interpretation":"His brief answer concerned what excited him in 2025. That calendar year has elapsed, but this was an expression of expectation rather than a technical commitment.","confidence":"Brief expectation, not a detailed forecast"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2024-20vc","url":"https://whocriedagi.com/claims/altman-2024-20vc","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2024-11-04","datePrecision":"day","quote":"Within five years, people may look back on an AGI milestone that has already passed.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He expected rapid technical progress while the immediate effect on society could be smaller than people imagined.","source":"https://www.youtube.com/watch?v=peg-aX1oii4","sourceLabel":"20VC / Harry Stebbings","evidence":"Interview published November 4; corroborated by Windows Central's November 5 report.","target":"2029","expectedArrival":{"sourceDate":"2024-11-04","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"Looking back five years later on an AGI milestone implies arrival by approximately 2029, not only in 2029.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2024-moonshots","url":"https://whocriedagi.com/claims/kurzweil-2024-moonshots","name":"Ray Kurzweil","role":"Principal researcher and futurist","company":"Google","date":"2024-10-24","datePrecision":"day","quote":"AGI by 2029 is conservative.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In a Moonshots interview with Peter Diamandis, Kurzweil again defended 2029 as a cautious estimate.","source":"https://www.youtube.com/watch?v=sk1gfhAf6kU","sourceLabel":"Moonshots / Peter Diamandis","evidence":"Official interview published October 24; wording appears in its description.","target":"2029","expectedArrival":{"sourceDate":"2024-10-24","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"The claim is explicitly by 2029; no lower arrival bound is implied."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"son-2024-softbank-world","url":"https://whocriedagi.com/claims/son-2024-softbank-world","name":"Masayoshi Son","role":"Chairman and CEO","company":"SoftBank Group","date":"2024-10-03","datePrecision":"day","quote":"AGI will be achieved in two to three years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At SoftBank World, Son shortened his earlier horizon and separately predicted superintelligence within a decade.","source":"https://www.marketscreener.com/quote/stock/SOFTBANK-GROUP-CORP-6492452/news/SoftBank-Group-CEO-Says-Artificial-General-Intelligence-to-Be-Achieved-in-Two-to-Three-Years-47989311/","sourceLabel":"Dow Jones / SoftBank World","evidence":"Contemporaneous October 3 report; also summarized by SoftBank's corporate account.","target":"2026–2027","expectedArrival":{"sourceDate":"2024-10-03","sourceDatePrecision":"day","kind":"range","startYear":2026,"endYear":2027,"classification":"unspecified","interpretation":"Two to three years from October 2024 gives a nominal October 2026–October 2027 horizon. It overlaps the current year, but no precise day-level boundary was promised.","approximate":true,"currentYearOverlap":true,"ambiguity":"Today is near the approximate lower boundary; exact overlap is not established."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2024-machines-loving-grace","url":"https://whocriedagi.com/claims/amodei-2024-machines-loving-grace","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2024-10-01","datePrecision":"month","quote":"I think it could come as early as 2026","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He used 'powerful AI' for systems exceeding leading specialists across fields and acting autonomously over extended periods. He said the forecast was uncertain and could take much longer.","source":"https://darioamodei.com/essay/machines-of-loving-grace","sourceLabel":"Machines of Loving Grace","evidence":"Amodei's essay is dated October 2024.","target":"2026 at the earliest","expectedArrival":{"sourceDate":"2024-10-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"2026 is an earliest plausible marker, not an upper deadline. The explicit possibility of a much later arrival prevents a bounded window.","earliestPossibleYear":2026},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"lecun-2024-hudson-world-models","url":"https://whocriedagi.com/claims/lecun-2024-hudson-world-models","name":"Yann LeCun","role":"Chief AI Scientist","company":"Meta","date":"2024-10-01","datePrecision":"month","quote":"Human-level AI remains years to decades away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Requires world models.","source":"https://techcrunch.com/2024/10/16/metas-ai-chief-says-world-models-are-key-to-human-level-ai-but-it-might-be-10-years-out/","sourceLabel":"TechCrunch","evidence":"Hudson Forum remarks.","target":"Years to decades","expectedArrival":{"sourceDate":"2024-10-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"Years to decades does not provide numerical bounds. No particular count of years or decades is inferred."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2024-intelligence-age","url":"https://whocriedagi.com/claims/altman-2024-intelligence-age","name":"Sam Altman","role":"Co-founder and CEO","company":"OpenAI","date":"2024-09-23","datePrecision":"day","quote":"It is possible that we will have superintelligence in a few thousand days","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"Altman explicitly allowed that it could take longer. This forecast concerns superintelligence, a higher bar than many definitions of AGI.","source":"https://ia.samaltman.com/","sourceLabel":"The Intelligence Age","evidence":"Altman's essay dated September 23, 2024.","target":"A few thousand days, possibly longer","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2024-time-impact","url":"https://whocriedagi.com/claims/kurzweil-2024-time-impact","name":"Ray Kurzweil","role":"Principal researcher and futurist","company":"Google","date":"2024-09-16","datePrecision":"day","quote":"AGI, matching or surpassing human intelligence generally, will arrive by 2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He reiterated his timeline at TIME's AI Impact Dinner in San Francisco.","source":"https://time.com/7021119/time100-impact-dinner-ai-transformative-potential-kurzweil-okolo-kelly-kumar/","sourceLabel":"TIME100 Impact Dinner","evidence":"TIME's September 17 report identifies the event as Monday.","target":"2029","expectedArrival":{"sourceDate":"2024-09-16","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"An explicit by-2029 prediction, not a closed arrival interval."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2024-time100","url":"https://whocriedagi.com/claims/hassabis-2024-time100","name":"Demis Hassabis","role":"Co-founder and CEO","company":"Google DeepMind","date":"2024-09-05","datePrecision":"day","quote":"DeepMind remains on track for the twenty-year AGI plan it began in 2010.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The profile tied his progress assessment to a roughly 2030 horizon. He still expected further breakthroughs before genuinely general intelligence.","source":"https://time.com/7012767/demis-hassabis/","sourceLabel":"TIME100 AI","evidence":"TIME profile published September 5.","target":"Around 2030","expectedArrival":{"sourceDate":"2024-09-05","sourceDatePrecision":"day","kind":"around","targetYear":2030,"classification":"unspecified","interpretation":"A twenty-year research plan beginning in 2010 supplies a nominal 2030 milestone, not a measured arrival window.","estimateType":"plan-horizon","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1825577151329878291","url":"https://whocriedagi.com/claims/tweet-sweep-x-1825577151329878291","name":"Matt Brooks","role":"Commentator · @qwweryo","company":"","date":"2024-08-19","datePrecision":"day","quote":"And it seems like full blown agentic AGI by 2050 is quite likely","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Brooks calls agentic AGI by 2050 quite likely. “Quite likely” remains qualitative; no numerical probability is inferred.","source":"https://x.com/qwweryo/status/1825577151329878291","sourceLabel":"Matt Brooks on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2024-08-19","kind":"by","deadlineYear":2050,"interpretation":"Agentic AGI by 2050 is described as quite likely."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2024-guardian","url":"https://whocriedagi.com/claims/kurzweil-2024-guardian","name":"Ray Kurzweil","role":"Principal researcher and futurist","company":"Google","date":"2024-06-29","datePrecision":"day","quote":"Human-level AGI will arrive by 2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Reaffirmed his longstanding forecast.","source":"https://www.theguardian.com/technology/article/2024/jun/29/ray-kurzweil-google-ai-the-singularity-is-nearer","sourceLabel":"The Guardian","evidence":"Published interview.","target":"2029","expectedArrival":{"sourceDate":"2024-06-29","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"A reaffirmed by-2029 forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"son-2024-annual-message","url":"https://whocriedagi.com/claims/son-2024-annual-message","name":"Masayoshi Son","role":"Chairman and CEO","company":"SoftBank Group","date":"2024-06-21","datePrecision":"day","quote":"In less than ten years, AGI will be realized","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"His annual-report message anticipated AGI first, then vastly more capable superintelligence in approximately ten years. These are separate capability thresholds.","source":"https://group.softbank/en/ir/financials/annual_reports/2024/message/son","sourceLabel":"SoftBank Group annual report","evidence":"Son's signed message is dated June 21, 2024.","target":"Before 2034","expectedArrival":{"sourceDate":"2024-06-21","sourceDatePrecision":"day","kind":"by","deadlineYear":2034,"classification":"window-open","interpretation":"Less than ten years from June 2024 supplies an outer horizon before the 2034 anniversary. No lower bound is stated.","deadlineRelation":"before-relative-anniversary"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"yampolskiy-2024-lex-average-human","url":"https://whocriedagi.com/claims/yampolskiy-2024-lex-average-human","name":"Roman Yampolskiy","role":"AI safety researcher","company":"","date":"2024-06-02","datePrecision":"day","quote":"Current systems already qualified as AGI if the threshold was average performance across common human tasks.","paraphrase":true,"kind":"qualified","topic":"agi","models":[],"context":"He distinguished this average-human interpretation from outperforming elite humans in every domain, which he said had not happened. No particular model was named in this passage.","source":"https://lexfridman.com/roman-yampolskiy-transcript/","sourceLabel":"Lex Fridman — original interview transcript","evidence":"Transcript 00:20:55; June 2, 2024 publication verified on publisher episode archive.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"aschenbrenner-2024-situational-awareness","url":"https://whocriedagi.com/claims/aschenbrenner-2024-situational-awareness","name":"Leopold Aschenbrenner","role":"Former OpenAI researcher","company":"Independent","date":"2024-06-01","datePrecision":"month","quote":"AGI by 2027 is strikingly plausible.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"His essay extrapolated gains in compute, algorithmic efficiency and the ability to turn models into useful agents. It presented a scenario from public evidence, not inside information.","source":"https://situational-awareness.ai/","sourceLabel":"Situational Awareness","evidence":"Author's essay collection, dated June 2024.","target":"2027","expectedArrival":{"sourceDate":"2024-06-01","sourceDatePrecision":"month","kind":"by","deadlineYear":2027,"classification":"window-open","interpretation":"By 2027 is described as plausible, not as certain or assigned a numerical probability.","confidence":"Strikingly plausible; no numerical probability"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"schulman-2024-dwarkesh-timeline","url":"https://whocriedagi.com/claims/schulman-2024-dwarkesh-timeline","name":"John Schulman","role":"Cofounder","company":"OpenAI","date":"2024-05-15","datePrecision":"day","quote":"He did not expect AGI the following year, but considered its arrival in two or three years possible.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Responding to a hypothetical early arrival, he discussed coordination and slowing training or deployment. This was a possibility, not a definite 2027 deadline.","source":"https://www.dwarkesh.com/p/john-schulman","sourceLabel":"Dwarkesh Podcast — original interview transcript","evidence":"Plan for AGI section, 00:16:55; explicit response rejects the host's next-year assumption.","target":"2026–2027 possible","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1786795779035926887","url":"https://whocriedagi.com/claims/tweet-sweep-x-1786795779035926887","name":"Rob Haisfield","role":"Commentator","company":"","date":"2024-05-04","datePrecision":"day","quote":"AGI is already here and if you used Claude in Websim you’d know.","paraphrase":false,"kind":"arrived","topic":"agi","models":["Claude"],"context":"Haisfield recounts telling friends that AGI is already here, pointing to Claude in Websim. This records his own explicit arrival claim; the post does not specify a Claude version or an operational AGI definition.","source":"https://x.com/RobertHaisfield/status/1786795779035926887","sourceLabel":"Rob Haisfield on X","evidence":"Discovered in Community Archive’s current public export and independently checked against X’s official public oEmbed. Author, post ID, visible excerpt and UTC publication date are verified. The statement records the speaker’s view, not this archive’s assessment of AGI.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2024-tangen-interview","url":"https://whocriedagi.com/claims/musk-2024-tangen-interview","name":"Elon Musk","role":"Founder","company":"xAI","date":"2024-04-08","datePrecision":"day","quote":"I think it's probably next year, within two years","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Musk defined AGI as intelligence exceeding the smartest human. He gave this estimate in an X Spaces interview with Nicolai Tangen.","source":"https://finance.yahoo.com/news/1-teslas-musk-predicts-ai-190932582.html","sourceLabel":"Reuters / Nicolai Tangen interview","evidence":"Reuters April 8 report quotes the interview.","target":"2025–2026","expectedArrival":{"sourceDate":"2024-04-08","sourceDatePrecision":"day","kind":"by","deadlineYear":2026,"classification":"window-elapsed","interpretation":"The outer limit was within two years of April 8, 2024, with 2025 described as probable. The nominal two-year horizon has elapsed by October 2026.","deadlineMonth":4,"deadlineBasis":"Two-year relative horizon anchored to the dated interview","preferredYear":2025,"confidence":"Probably next year; outer two-year estimate"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1771047845610639506","url":"https://whocriedagi.com/claims/tweet-sweep-x-1771047845610639506","name":"roon","role":"Commentator · @tszzl","company":"","date":"2024-03-22","datePrecision":"day","quote":"GPT3 was agi","paraphrase":false,"kind":"arrived","topic":"agi","models":["GPT-3"],"context":"Roon retrospectively calls GPT-3 AGI in March 2024. The short statement gives neither an operational definition nor a date when the speaker first held this view; it remains dated to publication.","source":"https://x.com/tszzl/status/1771047845610639506","sourceLabel":"roon on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1766820971946647695","url":"https://whocriedagi.com/claims/tweet-sweep-x-1766820971946647695","name":"David R. MacIver","role":"Commentator · @DRMacIver","company":"","date":"2024-03-10","datePrecision":"day","quote":"MacIver would be extremely surprised if AGI had not arrived within fifty years, and expects some connection to the current generation of chatbots.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"MacIver expects AGI within fifty years of March 2024, potentially descended from current chatbots even if only through their training data. He does not give a probability or a nearer central estimate.","source":"https://x.com/DRMacIver/status/1766820971946647695","sourceLabel":"David R. MacIver on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2024-03-10","kind":"by","deadlineYear":2074,"approximate":true,"interpretation":"AGI within fifty years of March 2024."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2024-beneficial-agi-summit","url":"https://whocriedagi.com/claims/goertzel-2024-beneficial-agi-summit","name":"Ben Goertzel","role":"Founder and CEO","company":"SingularityNET","date":"2024-03-01","datePrecision":"day","quote":"Human-level or superhuman AGI is likely by 2029–2030, and could arrive as early as 2027.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"In closing remarks at the Beneficial AGI Summit, he described 2027 as a possibility rather than a firm deadline.","source":"https://futurism.com/artificial-superintelligence-agi-2027-goertzel","sourceLabel":"Futurism / Beneficial AGI Summit","evidence":"March 7 report of the summit's March 1 closing remarks.","target":"2027 possible; 2029–2030 more likely","expectedArrival":{"sourceDate":"2024-03-01","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"The likely by-2029–2030 horizon is represented by its outer deadline; 2027 is separately an early possibility. No 2027–2030 closed interval is implied.","deadlineEstimateStartYear":2029,"earliestPossibleYear":2027,"approximate":true,"confidence":"2029–2030 more likely; 2027 possible"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"huang-2024-stanford-five-years","url":"https://whocriedagi.com/claims/huang-2024-stanford-five-years","name":"Jensen Huang","role":"Co-founder and CEO","company":"NVIDIA","date":"2024-03-01","datePrecision":"day","quote":"AI could pass every human test within five years, if that is the definition of AGI.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At Stanford, Huang made his five-year prediction conditional on defining AGI through human tests. He distinguished this from reproducing how the brain works.","source":"https://www.investing.com/news/stock-market-news/nvidia-ceo-says-ai-could-pass-human-tests-in-five-years-3322301","sourceLabel":"Reuters / Stanford SIEPR","evidence":"Reuters report of his March 1 Stanford appearance.","target":"2029, under an exam-based definition","expectedArrival":{"sourceDate":"2024-03-01","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"Within five years of March 2024 is an outer horizon, conditional on an exam-based AGI definition.","approximate":true,"condition":"AGI defined by passing human tests"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1762559116382187996","url":"https://whocriedagi.com/claims/tweet-sweep-x-1762559116382187996","name":"Aaron Bergman","role":"Commentator · @AaronBergman18","company":"","date":"2024-02-27","datePrecision":"day","quote":"Bergman gives a 50% chance of AGI by 2027, while emphasizing uncertainty and presenting several other probability thresholds.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The post gives a distribution rather than one confident date: 20% by the end of 2024, 40% by 2026, 50% by 2027, 70% by 2030, 80% by 2035, 90% by 2040, and 99% before 2050 conditional on no nuclear war. Bergman says regulation adds uncertainty and cautions readers to take his estimates with a grain of salt. The plotted horizon uses his stated 50% “actual guess.”","source":"https://x.com/AaronBergman18/status/1762559116382187996","sourceLabel":"Aaron Bergman on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2024-02-27","kind":"by","deadlineYear":2027,"probability":{"value":0.5},"interpretation":"50% chance by 2027, within a broader and explicitly tentative probability distribution."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"ng-2024-decades-away","url":"https://whocriedagi.com/claims/ng-2024-decades-away","name":"Andrew Ng","role":"Founder","company":"DeepLearning.AI","date":"2024-01-17","datePrecision":"day","quote":"AGI capable of any human intellectual task remained at least decades away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He warned that lowering AGI's definition could produce premature declarations that misled the public.","source":"https://www.deeplearning.ai/the-batch/the-easiest-way-to-achieve-artificial-general-intelligence","sourceLabel":"The Batch — Andrew Ng's letter","evidence":"Author-signed letter dated January 17, 2024; explicit decades-away judgment.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1740641473849352450","url":"https://whocriedagi.com/claims/social-followup-x-1740641473849352450","name":"Elon Musk","role":"CEO","company":"Tesla","date":"2023-12-29","datePrecision":"day","quote":"It is increasingly clear that all roads lead to AGI. Tesla is building an extremely compute-efficient mini AGI for FSD.","paraphrase":false,"kind":"qualified","topic":"agi","models":["FSD"],"context":"Musk applies the term “mini AGI” to Tesla’s in-development self-driving work. The qualifier and ongoing development matter: this is not an assertion that Tesla has completed a generally capable AGI. No arrival date is supplied.","source":"https://x.com/elonmusk/status/1740641473849352450","sourceLabel":"Elon Musk on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","originalTerminology":"mini AGI","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2023-time","url":"https://whocriedagi.com/claims/altman-2023-time","name":"Sam Altman","role":"CEO","company":"OpenAI","date":"2023-12-06","datePrecision":"day","quote":"AGI could be reached in the next four or five years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"TIME reported his expectation in its CEO of the Year profile. The statement describes a forecast, not a released model.","source":"https://time.com/6342827/ceo-of-the-year-2023-sam-altman/","sourceLabel":"TIME — CEO of the Year","evidence":"Original reported profile published December 6, 2023; publication date used.","target":"2027–2028","expectedArrival":{"sourceDate":"2023-12-06","sourceDatePrecision":"day","kind":"range","startYear":2027,"endYear":2028,"classification":"ahead","interpretation":"Four or five years from 2023 gives an approximate 2027–2028 horizon.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"huang-2023-dealbook","url":"https://whocriedagi.com/claims/huang-2023-dealbook","name":"Jensen Huang","role":"Founder and CEO","company":"Nvidia","date":"2023-11-29","datePrecision":"day","quote":"AI could match ordinary humans on a broad set of intelligence tests within five years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His forecast explicitly depended on defining AGI through a collection of human intelligence tests. He acknowledged unresolved multistep reasoning.","source":"https://www.youtube.com/watch?v=Pkj-BLHs6dE","sourceLabel":"New York Times DealBook Summit","evidence":"Original NYT Events interview; summit November 29, video published November 30; transcript cross-checked.","target":"By 2028","expectedArrival":{"sourceDate":"2023-11-29","sourceDatePrecision":"day","kind":"by","deadlineYear":2028,"classification":"window-open","interpretation":"Within five years of November 2023; conditional on a broad human-test definition, not a lower-and-upper arrival band.","approximate":true,"condition":"AGI defined through human intelligence tests"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2023-decrypt","url":"https://whocriedagi.com/claims/goertzel-2023-decrypt","name":"Ben Goertzel","role":"Founder and CEO","company":"SingularityNET","date":"2023-11-06","datePrecision":"day","quote":"I would say now, three to eight years is my take","paraphrase":false,"kind":"forecast","topic":"agi","models":["GPT-4","Llama 2"],"context":"He credited GPT-4 and Llama 2 with progress and increased enthusiasm, funding, and research talent for AGI. He forecast future AGI rather than declaring those models had achieved it.","source":"https://decrypt.co/204571","sourceLabel":"Decrypt interview","evidence":"Original interview published November 6, 2023.","target":"2026–2031","expectedArrival":{"sourceDate":"2023-11-06","sourceDatePrecision":"day","kind":"range","startYear":2026,"endYear":2031,"classification":"unspecified","interpretation":"Three to eight years from the November 2023 publication gives a nominal late-2026–2031 horizon. The exact interview day and within-year lower boundary are not established.","approximate":true,"currentYearOverlap":true,"ambiguity":"The nominal lower year is 2026; exact overlap on October 7 is uncertain."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2023-dwarkesh","url":"https://whocriedagi.com/claims/legg-2023-dwarkesh","name":"Shane Legg","role":"Co-founder and chief AGI scientist","company":"Google DeepMind","date":"2023-10-26","datePrecision":"day","quote":"I think there's a 50% chance that we have AGI by 2028.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He reaffirmed a long-held forecast and stressed that missing the date would be unsurprising. His definition covered the cognitive tasks humans typically perform.","source":"https://www.dwarkesh.com/p/shane-legg","sourceLabel":"Dwarkesh Podcast","evidence":"Original transcript, 34:38–37:13; published October 26, 2023.","target":"2028 (50%)","expectedArrival":{"sourceDate":"2023-10-26","sourceDatePrecision":"day","kind":"by","deadlineYear":2028,"classification":"window-open","interpretation":"A cumulative 50% chance by 2028, not certainty or an isolated arrival year.","probability":{"value":0.5,"scope":"by-deadline","approximate":true}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"aguera-noema-2023","url":"https://whocriedagi.com/claims/aguera-noema-2023","name":"Blaise Agüera y Arcas","role":"VP & Fellow · with Peter Norvig","company":"Google Research","date":"2023-10-10","datePrecision":"day","quote":"Artificial General Intelligence Is Already Here","paraphrase":false,"kind":"arrived","topic":"agi","models":["ChatGPT","Bard","LLaMA","Claude"],"context":"In a co-authored essay, Agüera y Arcas and Peter Norvig argued that the generality of contemporary language models already satisfies key aspects of AGI. They named multiple systems rather than a single winner.","source":"https://www.noemamag.com/artificial-general-intelligence-is-already-here/","sourceLabel":"Noema · co-authored essay","evidence":"Primary essay. The displayed excerpt is its title; the argument uses a broad definition of general intelligence.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"son-2023-softbank-world","url":"https://whocriedagi.com/claims/son-2023-softbank-world","name":"Masayoshi Son","role":"Chairman and CEO","company":"SoftBank Group","date":"2023-10-04","datePrecision":"day","quote":"AGI will surpass humanity's combined intelligence tenfold within ten years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Son used an unusually strong definition of AGI: exceeding humans across almost every field. This is an English paraphrase of his Japanese keynote.","source":"https://www.softbank.jp/business/content/blog/202310/sbw2023-softbank-son-main-keynote","sourceLabel":"SoftBank World — official keynote report","evidence":"Company transcript of the October 4 speech, initially published that day and updated October 12.","target":"By 2033","expectedArrival":{"sourceDate":"2023-10-04","sourceDatePrecision":"day","kind":"by","deadlineYear":2033,"classification":"window-open","interpretation":"Within ten years of the 2023 keynote; no lower bound is inferred."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"carmack-2023-keen","url":"https://whocriedagi.com/claims/carmack-2023-keen","name":"John Carmack","role":"Founder","company":"Keen Technologies","date":"2023-09-25","datePrecision":"day","quote":"We are six, seven, eight years out from something really big and important being publicly visible.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"At Richard Sutton's hiring announcement, Carmack forecast visible AGI progress while admitting nobody yet had a complete solution.","source":"https://www.theregister.com/software/2023/09/26/doom-developer-john-carmack-thinks-agi-could-debut-in-2030/437426","sourceLabel":"The Register — Keen announcement","evidence":"September 26 report identifies the announcement as Monday, September 25.","target":"2029–2031","expectedArrival":{"sourceDate":"2023-09-25","sourceDatePrecision":"day","kind":"range","startYear":2029,"endYear":2031,"classification":"ahead","interpretation":"Six, seven or eight years from 2023 gives this approximate horizon for publicly visible AGI progress.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"sutton-2023-keen","url":"https://whocriedagi.com/claims/sutton-2023-keen","name":"Richard Sutton","role":"Research scientist; chief scientific adviser","company":"Keen Technologies / Amii","date":"2023-09-25","datePrecision":"day","quote":"A prototype AGI showing early signs of life by 2030 is possible.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He thought existing techniques could support the effort, while Keen still faced fundamental research questions.","source":"https://www.theregister.com/software/2023/09/26/doom-developer-john-carmack-thinks-agi-could-debut-in-2030/437426","sourceLabel":"The Register — Keen announcement","evidence":"Statement at the September 25 event announcing Sutton's appointment.","target":"2030","expectedArrival":{"sourceDate":"2023-09-25","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"A possible prototype by 2030. The weaker prototype milestone is preserved.","confidence":"Possible","milestone":"Prototype AGI / early signs of life"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1695071731566662057","url":"https://whocriedagi.com/claims/tweet-sweep-x-1695071731566662057","name":"Eli Lifland","role":"Forecaster","company":"","date":"2023-08-25","datePrecision":"day","quote":"Would never have guessed Gary Marcus was at 35% for AGI within 13 years, that's pretty close to my own views","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Lifland says a reported 35% probability of AGI within thirteen years is close to his own view. The record captures Lifland’s endorsement, not an independently verified Gary Marcus forecast. His own probability is approximate rather than exactly 35%.","source":"https://x.com/eli_lifland/status/1695071731566662057","sourceLabel":"Eli Lifland on X","evidence":"Discovered in Community Archive’s current public export and independently checked against X’s official public oEmbed. Author, post ID, visible excerpt and UTC publication date are verified. The statement records the speaker’s view, not this archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2023-08-25","kind":"by","deadlineYear":2036,"approximate":true,"interpretation":"Approximately 35% within thirteen years of August 2023.","note":"Lifland says the cited estimate is close to his own; it is not his exact numerical forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"amodei-2023-dwarkesh","url":"https://whocriedagi.com/claims/amodei-2023-dwarkesh","name":"Dario Amodei","role":"Co-founder and CEO","company":"Anthropic","date":"2023-08-08","datePrecision":"day","quote":"A model resembling a generally educated human in conversation could be two or three years away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He separated this conversational threshold from replacing researchers, transforming the economy or becoming existentially dangerous. Safety restrictions could slow the forecast.","source":"https://www.dwarkesh.com/p/dario-amodei","sourceLabel":"Dwarkesh Podcast","evidence":"Original transcript, 27:56–29:32; published August 8, 2023.","target":"2025–2026","expectedArrival":{"sourceDate":"2023-08-08","sourceDatePrecision":"day","kind":"range","startYear":2025,"endYear":2026,"classification":"unspecified","interpretation":"Two or three years from the August 2023 interview gives a nominal August 2025–August 2026 estimate for human-like conversation. Safety restrictions could delay it; no firm within-year cutoff was promised.","approximate":true,"currentYearOverlap":true,"ambiguity":"Nominal anniversary has passed, but the approximate horizon does not support a precise October 7 verdict.","evidence":"Primary transcript rechecked at 27:56–29:32; this is a conversational threshold, not full economic substitution.","source":"https://www.dwarkesh.com/p/dario-amodei"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"bengio-2023-senate","url":"https://whocriedagi.com/claims/bengio-2023-senate","name":"Yoshua Bengio","role":"Founder and scientific director","company":"Mila","date":"2023-07-25","datePrecision":"day","quote":"Human-level AI could arrive in the next two decades, possibly within a few years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His written Senate testimony framed this as a possibility motivating precautions, not a declaration of arrival.","source":"https://www.judiciary.senate.gov/imo/media/doc/2023-07-26_-_testimony_-_bengio.pdf","sourceLabel":"U.S. Senate written testimony","evidence":"Original Senate-hosted testimony dated July 25, 2023 on its cover; the filename's July 26 is not the hearing date.","target":"Possibly a few years; within two decades","expectedArrival":{"sourceDate":"2023-07-25","sourceDatePrecision":"day","kind":"by","deadlineYear":2043,"classification":"window-open","interpretation":"Within two decades from 2023 supplies an outer possible horizon. The separate few-years possibility has no numerical lower bound.","approximate":true,"confidence":"Possible, not a guaranteed deadline"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2023-xai","url":"https://whocriedagi.com/claims/musk-2023-xai","name":"Elon Musk","role":"Founder and CEO","company":"xAI / Tesla / SpaceX","date":"2023-07-12","datePrecision":"day","quote":"Digital superintelligence could arrive in five or six years.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"In a Twitter Spaces discussion introducing xAI, Musk forecast superhuman AI. He did not claim the newly announced company already had such a model.","source":"https://www.abc.net.au/news/2023-07-13/elon-musk-launches-ai-firm-xai-as-he-looks-to-take-on-openai/102599786","sourceLabel":"ABC / Reuters on xAI launch","evidence":"Contemporaneous July 13 report identifies the discussion as Wednesday evening, July 12.","target":"2028–2029","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"sutskever-leike-2023-superalignment","url":"https://whocriedagi.com/claims/sutskever-leike-2023-superalignment","name":"Ilya Sutskever & Jan Leike","role":"Chief scientist and head of alignment","company":"OpenAI","date":"2023-07-05","datePrecision":"day","quote":"superintelligence could arrive this decade","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"The team announcement explicitly targeted capabilities beyond AGI and said the timing remained uncertain. Its four-year research deadline was not an AGI release promise.","source":"https://openai.com/index/introducing-superalignment/","sourceLabel":"OpenAI — Introducing Superalignment","evidence":"Original announcement, July 5, 2023.","target":"Before 2030","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"bengio-2023-faq","url":"https://whocriedagi.com/claims/bengio-2023-faq","name":"Yoshua Bengio","role":"Founder and scientific director","company":"Mila","date":"2023-06-24","datePrecision":"day","quote":"My current estimate places a 95% confidence interval for the time horizon of superhuman intelligence at 5 to 20 years.","paraphrase":false,"kind":"forecast","topic":"asi","models":["ChatGPT","GPT-4"],"context":"Bengio explicitly reduced his earlier 20–100-year horizon after these systems' progress. He still identified missing capabilities and uncertainty.","source":"https://yoshuabengio.org/en/blog/faq-catastrophic-ai-risks","sourceLabel":"Yoshua Bengio — personal essay","evidence":"Original essay dated June 24; answers A1 and A3a.","target":"2028–2043 (95% interval)","expectedArrival":{"sourceDate":"2023-06-24","sourceDatePrecision":"day","kind":"range","startYear":2028,"endYear":2043,"classification":"ahead","interpretation":"His stated 95% interval spans five to twenty years from 2023. This is an uncertainty interval, not a claim that outcomes outside it are impossible.","confidenceInterval":0.95,"approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-brockman-sutskever-2023-governance","url":"https://whocriedagi.com/claims/altman-brockman-sutskever-2023-governance","name":"Sam Altman, Greg Brockman & Ilya Sutskever","role":"CEO, president and chief scientist","company":"OpenAI","date":"2023-05-22","datePrecision":"day","quote":"AI exceeding expert ability in most domains could arrive within ten years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Their joint essay described superintelligence as beyond AGI, potentially doing as much productive work as a large corporation. They used possibility language.","source":"https://openai.com/index/governance-of-superintelligence/","sourceLabel":"OpenAI — Governance of superintelligence","evidence":"Original joint essay, May 22, 2023.","target":"By 2033","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"davidson-2023-80000-takeoff","url":"https://whocriedagi.com/claims/davidson-2023-80000-takeoff","name":"Tom Davidson","role":"Senior research analyst","company":"Open Philanthropy","date":"2023-05-05","datePrecision":"day","quote":"If AI became highly profitable or substantially accelerated AI research this decade, it was hard to see AGI remaining absent in 2040.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He expected feedback from investment and AI-assisted research to bring AGI sooner than a simple compute extrapolation suggested. This remained conditional on AGI not being exceptionally difficult.","source":"https://80000hours.org/podcast/episodes/tom-davidson-how-quickly-ai-could-transform-the-world/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: Fast AI takeoff speeds probably mean shorter AI timelines, 02:34:44.","target":"2040, conditional","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1653687894534504451","url":"https://whocriedagi.com/claims/social-followup-x-1653687894534504451","name":"Geoffrey Hinton","role":"AI researcher","company":"University of Toronto","date":"2023-05-03","datePrecision":"day","quote":"I now predict 5 to 20 years but without much confidence.","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"Hinton explicitly says he has little confidence in this estimate and could be entirely wrong. The forecast concerns digital intelligence overtaking humans; it does not declare that an existing system is AGI.","source":"https://x.com/geoffreyhinton/status/1653687894534504451","sourceLabel":"Geoffrey Hinton on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","originalTerminology":"digital intelligence overtaking us","expectedArrival":{"kind":"range","sourceDate":"2023-05-03","startYear":2028,"endYear":2043,"interpretation":"Five to twenty years from May 2023.","note":"Hinton explicitly expresses low confidence and says he may be wrong."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2023-wsj","url":"https://whocriedagi.com/claims/hassabis-2023-wsj","name":"Demis Hassabis","role":"CEO and co-founder","company":"Google DeepMind","date":"2023-05-02","datePrecision":"day","quote":"Very capable, very general AI could arrive within a few years, perhaps a decade.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At the WSJ Future of Everything Festival, Hassabis expected AI progress to continue and potentially accelerate. His forecast spanned a few years to a decade.","source":"https://indianexpress.com/article/technology/artificial-intelligence/ai-as-powerful-as-human-mind-possible-in-next-5-years-demis-hassabis-8589820/","sourceLabel":"Indian Express on WSJ interview","evidence":"May 3 contemporaneous coverage of the May 2 interview; broader range preserved.","target":"A few years to 2033","expectedArrival":{"sourceDate":"2023-05-02","sourceDatePrecision":"day","kind":"by","deadlineYear":2033,"classification":"window-open","interpretation":"Perhaps a decade gives an approximate outer horizon. A few years remains vague and is not converted to a numerical starting year.","approximate":true,"confidence":"Tentative upper horizon"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2023-afp","url":"https://whocriedagi.com/claims/goertzel-2023-afp","name":"Ben Goertzel","role":"Founder and CEO","company":"SingularityNET","date":"2023-05-01","datePrecision":"month","quote":"there's reason to believe we're years rather than decades from getting there","paraphrase":false,"kind":"forecast","topic":"agi","models":["ChatGPT"],"context":"Asked about human cognitive abilities, he said AI still needed to generalize beyond its training. He explicitly said current ChatGPT-like systems lacked human-level general intelligence.","source":"https://techxplore.com/news/2023-05-ai-jobs-years-expert.html","sourceLabel":"AFP interview at Web Summit Rio","evidence":"Original AFP Q&A published May 9; interview occurred at the preceding week's summit, exact day unspecified.","target":"Years, not decades","expectedArrival":{"sourceDate":"2023-05-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"Years rather than decades is qualitative; no numerical interval or deadline is stated."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1650497721948577792","url":"https://whocriedagi.com/claims/tweet-sweep-x-1650497721948577792","name":"Nathan Young","role":"Forecaster","company":"","date":"2023-04-24","datePrecision":"day","quote":"I think there is a 10% chance we have human level AGI in 5 years.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Young gives a 10% chance of human-level AGI within five years of April 2023. This low-probability near-term forecast is distinct from his longer-horizon forecasts.","source":"https://x.com/NathanpmYoung/status/1650497721948577792","sourceLabel":"Nathan Young on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2023-04-24","kind":"by","deadlineYear":2028,"approximate":true,"probability":{"value":0.1},"interpretation":"10% chance within five years of April 2023."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-followup-x-1640217165822578688","url":"https://whocriedagi.com/claims/social-followup-x-1640217165822578688","name":"Siqi Chen","role":"Co-founder and CEO","company":"Runway (financial planning)","date":"2023-03-27","datePrecision":"day","quote":"i have been told that gpt5 is scheduled to complete training this december and that openai expects it to achieve agi.","paraphrase":false,"kind":"forecast","topic":"agi","models":["GPT-5"],"context":"Chen relays an unnamed-source claim about GPT-5 training and OpenAI’s expectations, then predicts that debate over whether it qualifies will follow. This is Chen’s public statement of a rumor, not an official OpenAI forecast. December refers to the rumored completion of training, not an independently promised AGI release date. His company is the financial-planning business Runway, not the video-generation company.","source":"https://x.com/blader/status/1640217165822578688","sourceLabel":"Siqi Chen on X","evidence":"Author and exact excerpt verified in X’s official public oEmbed response. UTC date decoded from the post ID and checked against the embed date. The excerpt preserves the source’s wording; additional context is summarized separately.","expectedArrival":{"sourceDate":"2023-03-27","kind":"unspecified","interpretation":"The model is predicted to achieve AGI, but December 2023 is stated as a training-completion schedule.","note":"No AGI delivery deadline is plotted from a rumored training date."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1640475064985350146","url":"https://whocriedagi.com/claims/tweet-sweep-x-1640475064985350146","name":"George","role":"Commentator · @georgejrjrjr","company":"","date":"2023-03-27","datePrecision":"day","quote":"By historical definitions, the simulcra running on GPT3.5/GPT4 are AGI --goalposts are flying.","paraphrase":false,"kind":"qualified","topic":"agi","models":["GPT-3.5","GPT-4"],"context":"The @georgejrjrjr account calls GPT-3.5 and GPT-4 AGI under historical definitions while distinguishing intelligence from persistent agency. The same post describes LLMs as text simulators that can host temporary agents, rather than agents in themselves.","source":"https://x.com/georgejrjrjr/status/1640475064985350146","sourceLabel":"George on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hinton-2023-cbs-march","url":"https://whocriedagi.com/claims/hinton-2023-cbs-march","name":"Geoffrey Hinton","role":"VP and engineering fellow; professor emeritus","company":"Google / University of Toronto","date":"2023-03-25","datePrecision":"day","quote":"Now I think it might be 20 years or less.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In a CBS interview, he shortened his general-purpose AI forecast from 20–50 years. This was not an assertion that ChatGPT had reached AGI.","source":"https://www.utoronto.ca/news/cbs-news-travels-toronto-explore-future-ai","sourceLabel":"University of Toronto on CBS interview","evidence":"University report reproduces the quotation; original CBS broadcast dated March 25, 2023.","target":"By 2043, possibly sooner","expectedArrival":{"sourceDate":"2023-03-25","sourceDatePrecision":"day","kind":"by","deadlineYear":2043,"classification":"window-open","interpretation":"Twenty years or less from 2023 is an upper horizon, allowing earlier arrival.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"bubeck-sparks-2023","url":"https://whocriedagi.com/claims/bubeck-sparks-2023","name":"Sébastien Bubeck","role":"Research lead · with co-authors","company":"Microsoft Research","date":"2023-03-22","datePrecision":"day","quote":"an early (yet still incomplete) version of an artificial general intelligence (AGI) system","paraphrase":false,"kind":"qualified","topic":"agi","models":["GPT-4"],"context":"Bubeck and co-authors studied an early GPT-4 model and argued its breadth of capabilities suggested an incomplete form of AGI. Their paper also discussed substantial limitations. This is a research interpretation, not a claim of completed AGI.","source":"https://arxiv.org/abs/2303.12712","sourceLabel":"Research paper · Sparks of AGI","evidence":"Primary research paper, first submitted March 22, 2023. Roles are recorded in the context of the statement, not as current employment claims.","expectedArrival":{"sourceDate":"2023-03-22","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"An interpretation of an existing model's incomplete generality, with no future dated forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1633473109184454657","url":"https://whocriedagi.com/claims/tweet-sweep-x-1633473109184454657","name":"Paul in NYC","role":"Commentator · @Paul_Melman","company":"","date":"2023-03-08","datePrecision":"day","quote":"AGI is here. It's not yet superhuman, but it is general intelligence","paraphrase":false,"kind":"qualified","topic":"agi","models":[],"context":"The @Paul_Melman account endorses AGI’s arrival while distinguishing general intelligence from superhuman performance. The post links to a paper attributed to Google; the visible statement does not name a particular model.","source":"https://x.com/Paul_Melman/status/1633473109184454657","sourceLabel":"Paul in NYC on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1627863250821074944","url":"https://whocriedagi.com/claims/tweet-sweep-x-1627863250821074944","name":"Liron Shapira","role":"Commentator","company":"","date":"2023-02-21","datePrecision":"day","quote":"Shapira expects a scenario involving superintelligent AGI and loss of human control most likely within a few years, rather than decades.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"In a thread about AI risk, Shapira endorses a near-term scenario involving superintelligent AGI. This is a forecast of a scenario, not evidence that AGI exists or that its outcome is inevitable.","source":"https://x.com/liron/status/1627863250821074944","sourceLabel":"Liron Shapira on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2023-02-21","kind":"unspecified","interpretation":"A few years after February 2023, without a numeric interval.","note":"Do not convert “a few” into an invented three- or five-year deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1625452035133915136","url":"https://whocriedagi.com/claims/tweet-sweep-x-1625452035133915136","name":"Rob Miles","role":"AI safety educator","company":"","date":"2023-02-14","datePrecision":"day","quote":"No AGI within next 100 years, absent some kind of civilisational decline/collapse, is super unlikely.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Miles considers a century without AGI very unlikely, conditional on avoiding major civilizational decline or collapse. This broad upper horizon does not imply that he expects AGI specifically in 2123, and he gives no numeric probability.","source":"https://x.com/robertskmiles/status/1625452035133915136","sourceLabel":"Rob Miles on X","evidence":"Discovered in Community Archive’s current public export and independently checked against X’s official public oEmbed. Author, post ID, visible excerpt and UTC publication date are verified. The statement records the speaker’s view, not this archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2023-02-14","kind":"by","deadlineYear":2123,"approximate":true,"interpretation":"Within a century of February 2023, conditional on avoiding civilizational decline or collapse.","note":"A broad conditional horizon, not a prediction specifically for the year 2123."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"social-marcus-2023-01-22-agi-will-not-happen-in-your-lifetime","url":"https://whocriedagi.com/claims/social-marcus-2023-01-22-agi-will-not-happen-in-your-lifetime","name":"Gary Marcus","role":"AI researcher and author","company":"Independent","date":"2023-01-22","datePrecision":"day","quote":"AGI may arrive before this century ends, but current language models lack several essential capabilities.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Marcus contrasts his own eventual-AGI optimism with Grady Booch’s expectation of a timescale spanning generations. The article is a conversation, so this record represents Marcus’s own position.","source":"https://garymarcus.substack.com/p/agi-will-not-happen-in-your-lifetime","sourceLabel":"Gary Marcus — AGI will not happen in your lifetime. Or will it?","evidence":"Full public author post and publication timestamp returned by the publication’s unauthenticated public API.","target":"Possibly before the end of the 21st century","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"carmack-2023-dallas","url":"https://whocriedagi.com/claims/carmack-2023-dallas","name":"John Carmack","role":"Founder","company":"Keen Technologies","date":"2023-01-01","datePrecision":"year","quote":"Maybe a 60% chance of initial AGI by 2030; about 95% by 2050.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He defined the first milestone as a computer-based being recognized as intelligent and conscious. He separated that from immediate large economic effects.","source":"https://dallasinnovates.com/exclusive-qa-john-carmacks-different-path-to-artificial-general-intelligence/","sourceLabel":"Dallas Innovates — original Q&A","evidence":"Published in the 2023 annual magazine and circulated online by February 6; exact interview/publication day unverified.","target":"2030 (60%); 2050 (95%)","expectedArrival":{"sourceDate":"2023-01-01","sourceDatePrecision":"year","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"The principal milestone is 60% by 2030. A separate 95% by 2050 estimate is preserved rather than merging them into a 2030–2050 range.","probability":{"value":0.6,"scope":"by-deadline","approximate":true},"probabilityMilestones":[{"deadlineYear":2030,"probability":0.6},{"deadlineYear":2050,"probability":0.95}]},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1591494093766787076","url":"https://whocriedagi.com/claims/tweet-sweep-x-1591494093766787076","name":"Joshua Achiam","role":"AI researcher","company":"","date":"2022-11-12","datePrecision":"day","quote":"My AGI timelines are short-ish (within two decades with things getting weird soon)","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Achiam forecasts AGI within two decades of November 2022 while rejecting highly confident near-term extinction estimates. This is his own timeline, not the separate doom probabilities he criticizes.","source":"https://x.com/jachiam0/status/1591494093766787076","sourceLabel":"Joshua Achiam on X","evidence":"Author, original post ID and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records the speaker’s forecast, not an assessment that AGI or ASI exists.","expectedArrival":{"sourceDate":"2022-11-12","kind":"by","deadlineYear":2042,"approximate":true,"interpretation":"Within two decades of November 2022.","note":"An approximate horizon; the source does not specify a calendar-day deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2022-lex","url":"https://whocriedagi.com/claims/kurzweil-2022-lex","name":"Ray Kurzweil","role":"Director of engineering; futurist","company":"Google","date":"2022-09-17","datePrecision":"day","quote":"AI should pass a rigorous Turing test by 2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He reaffirmed his established date and discussed extended questioning by knowledgeable judges, rather than a brief chatbot demonstration.","source":"https://lexfridman.com/ray-kurzweil/","sourceLabel":"Lex Fridman Podcast #321","evidence":"Original episode published September 17; 2029 claim also presented in the publisher's September 19 clip.","target":"2029","expectedArrival":{"sourceDate":"2022-09-17","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"A rigorous Turing-test forecast by 2029."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"carmack-2022-lex","url":"https://whocriedagi.com/claims/carmack-2022-lex","name":"John Carmack","role":"Founder; consulting CTO","company":"Keen Technologies / Meta","date":"2022-08-04","datePrecision":"day","quote":"A 55–60% chance of AGI showing signs of life by 2030.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He raised an earlier 50% estimate slightly. The milestone was an initial demonstration, not a claim that robust AGI already existed.","source":"https://lexfridman.com/john-carmack","sourceLabel":"Lex Fridman Podcast #309","evidence":"Episode date verified on original publisher page; wording cross-checked against episode transcript.","target":"2030 (55–60%)","expectedArrival":{"sourceDate":"2022-08-04","sourceDatePrecision":"day","kind":"by","deadlineYear":2030,"classification":"window-open","interpretation":"The estimate assigns 55–60% probability to an initial demonstration by 2030.","probability":{"minimum":0.55,"maximum":0.6,"scope":"by-deadline","approximate":true},"milestone":"Initial AGI signs of life"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1554846695434895361","url":"https://whocriedagi.com/claims/tweet-sweep-x-1554846695434895361","name":"Nathan Young","role":"Forecaster","company":"","date":"2022-08-03","datePrecision":"day","quote":"Like I think on balance we have AGI in 20 years.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Young discusses how having children fits with his expectation of AGI within twenty years. “On balance” indicates a personal expectation without a stated numerical probability.","source":"https://x.com/NathanpmYoung/status/1554846695434895361","sourceLabel":"Nathan Young on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2022-08-03","kind":"by","deadlineYear":2042,"approximate":true,"interpretation":"AGI within twenty years of August 2022."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1533577009762209792","url":"https://whocriedagi.com/claims/tweet-sweep-x-1533577009762209792","name":"roon","role":"Commentator · @tszzl","company":"","date":"2022-06-05","datePrecision":"day","quote":"agi is already here","paraphrase":false,"kind":"qualified","topic":"agi","models":["GitHub Copilot"],"context":"Roon points to GitHub Copilot’s expected effect on programming productivity and uses economic growth as an AGI criterion. The arrival claim is tied to that economic interpretation rather than a demonstrated general-purpose autonomous system.","source":"https://x.com/tszzl/status/1533577009762209792","sourceLabel":"roon on X","evidence":"Original post identity and visible wording checked against X’s official public oEmbed. UTC publication date decoded from the post ID. This records an attributed personal claim, not the archive’s conclusion that AGI exists.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2022-2029-tweet","url":"https://whocriedagi.com/claims/musk-2022-2029-tweet","name":"Elon Musk","role":"CEO","company":"Tesla / SpaceX","date":"2022-05-30","datePrecision":"day","quote":"I’d be surprised if we don’t have AGI by then.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"A reply to Jack Dorsey identified 2029 as the pivotal year. No specific AI model or technical criterion was named.","source":"https://twitter.com/elonmusk/status/1531328534169493506","sourceLabel":"Twitter post","evidence":"Original post ID and May 30 date reproduced in Gary Marcus's response and a DLR-hosted bibliography.","target":"2029","expectedArrival":{"sourceDate":"2022-05-30","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"A by-2029 claim; surprise at failure is not converted into a numerical probability."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"de-freitas-2022-gato","url":"https://whocriedagi.com/claims/de-freitas-2022-gato","name":"Nando de Freitas","role":"Research director","company":"DeepMind","date":"2022-05-14","datePrecision":"day","quote":"Solving these scaling challenges is what will deliver AGI.","paraphrase":false,"kind":"forecast","topic":"agi","models":["Gato"],"context":"Responding to coverage of Gato, he argued scale, memory, data and efficiency were the remaining challenges. He did not say the released Gato had already achieved human-level AGI.","source":"https://thenextweb.com/news/deepmind-researcher-claims-new-gato-ai-could-lead-to-agi-says-game-is-over","sourceLabel":"The Next Web — reproduced Twitter thread","evidence":"Contemporaneous report reproduces the May 14 thread in full.","expectedArrival":{"sourceDate":"2022-05-14","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"A claim about a technical path to AGI without a calendar forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"tweet-sweep-x-1346579038333186050","url":"https://whocriedagi.com/claims/tweet-sweep-x-1346579038333186050","name":"christian","role":"Commentator · @cxgonzalez","company":"","date":"2021-01-05","datePrecision":"day","quote":"AGI is coming in 15 years","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The @cxgonzalez account explicitly forecasts AGI fifteen years after January 2021. The short post does not state an AGI definition or probability.","source":"https://x.com/cxgonzalez/status/1346579038333186050","sourceLabel":"christian on X","evidence":"Discovered through public archives and checked against X’s official public oEmbed. Author, post ID and visible wording verified; UTC publication date decoded from the post ID. This records the speaker’s view, not the archive’s assessment of AGI.","expectedArrival":{"sourceDate":"2021-01-05","kind":"around","targetYear":2036,"approximate":true,"interpretation":"Fifteen years after January 2021."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"musk-2020-nyt","url":"https://whocriedagi.com/claims/musk-2020-nyt","name":"Elon Musk","role":"CEO","company":"Tesla / SpaceX","date":"2020-07-25","datePrecision":"day","quote":"AI vastly smarter than humans could arrive in under five years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Forecast made in Maureen Dowd's interview, published July 25. It concerned future AI capability, with no named model.","source":"https://www.vanityfair.com/style/2020/07/elon-musk-on-the-ai-apocalypse-being-the-emperor-of-mars-and-texting-with-kanye-west","sourceLabel":"Vanity Fair on New York Times interview","evidence":"Contemporaneous July 25 report quoting Dowd's interview.","target":"Before July 2025","expectedArrival":{"sourceDate":"2020-07-25","sourceDatePrecision":"day","kind":"by","deadlineYear":2025,"classification":"window-elapsed","interpretation":"Under five years from July 2020 supplied a nominal deadline before July 2025, now elapsed. This classifies only the date, not whether the capability was achieved.","deadlineMonth":7,"deadlineRelation":"before-relative-anniversary"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"sutskever-2018-ai-frontiers","url":"https://whocriedagi.com/claims/sutskever-2018-ai-frontiers","name":"Ilya Sutskever","role":"Co-founder and chief scientist","company":"OpenAI","date":"2018-11-09","datePrecision":"day","quote":"near term AGI should be taken as a serious possibility","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"His conference presentation reviewed six years of AI progress. The statement expressed a possibility, with no specific deadline or probability.","source":"https://www.slideshare.net/slideshow/ilya-sutskever-at-ai-frontiers-progress-towards-the-openai-mission/123119210","sourceLabel":"AI Frontiers presentation","evidence":"Original slide deck dated November 9, 2018; also reported by Synced on November 13.","target":"Near term; unspecified","expectedArrival":{"sourceDate":"2018-11-09","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Near term is unquantified; no date is invented."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"christiano-2018-80000-takeoff","url":"https://whocriedagi.com/claims/christiano-2018-80000-takeoff","name":"Paul Christiano","role":"Researcher","company":"OpenAI","date":"2018-10-02","datePrecision":"day","quote":"A few years could separate major economic effects from AI doing everything, even under a comparatively gradual takeoff.","paraphrase":true,"kind":"forecast","topic":"both","models":[],"context":"He expected human-level AI to arrive in an already transformed world with many competing powerful systems. The two-year example was a plausible transition duration, not an absolute arrival deadline.","source":"https://80000hours.org/podcast/episodes/paul-christiano-ai-alignment-solutions/","sourceLabel":"80,000 Hours — original interview transcript","evidence":"Transcript: discussion of slow takeoff and human-level AI; speaker explicitly characterizes the scenario as speculative.","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2018-hlai","url":"https://whocriedagi.com/claims/goertzel-2018-hlai","name":"Ben Goertzel","role":"CEO; chief scientist","company":"SingularityNET / Hanson Robotics","date":"2018-08-01","datePrecision":"month","quote":"Human-level AI is less than ten years away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Conditional on existing algorithms sufficing when combined differently.","source":"https://www.weforum.org/stories/emerging-technologies/when-will-we-have-artificial-intelligence-as-smart-as-a-human-here-s-what-experts-think/","sourceLabel":"Futurism / World Economic Forum","evidence":"Conference interview; exact day unspecified.","target":"Before 2028","expectedArrival":{"sourceDate":"2018-08-01","sourceDatePrecision":"month","kind":"by","deadlineYear":2028,"classification":"window-open","interpretation":"Less than ten years from 2018 gives an outer 2028 horizon, conditional on the algorithmic approach succeeding.","deadlineRelation":"before-relative-anniversary","condition":"Existing algorithms suffice when combined differently"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kanai-2018-hlai","url":"https://whocriedagi.com/claims/kanai-2018-hlai","name":"Ryota Kanai","role":"CEO","company":"Araya","date":"2018-08-01","datePrecision":"month","quote":"Human-level AI in 2030.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"An explicitly speculative conference lightning-round guess.","source":"https://www.weforum.org/stories/emerging-technologies/when-will-we-have-artificial-intelligence-as-smart-as-a-human-here-s-what-experts-think/","sourceLabel":"Futurism / World Economic Forum","evidence":"Conference response; exact day unspecified.","target":"2030","expectedArrival":{"sourceDate":"2018-08-01","sourceDatePrecision":"month","kind":"range","startYear":2030,"endYear":2030,"classification":"ahead","interpretation":"The speculative answer names calendar year 2030, without a probability or finer timing.","confidence":"Speculative lightning-round guess"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kordik-2018-hlai","url":"https://whocriedagi.com/claims/kordik-2018-hlai","name":"Pavel Kordík","role":"Co-founder; associate professor","company":"Recombee / Czech Technical University","date":"2018-08-01","datePrecision":"month","quote":"Human-level AI might arrive in five years or later.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A speculative answer that stressed definition uncertainty.","source":"https://www.weforum.org/stories/emerging-technologies/when-will-we-have-artificial-intelligence-as-smart-as-a-human-here-s-what-experts-think/","sourceLabel":"Futurism / World Economic Forum","evidence":"Conference response; exact day unspecified.","target":"2023 or later","expectedArrival":{"sourceDate":"2018-08-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"Five years or later gives only an earliest-year marker of 2023. The open-ended later tail is not a missed deadline.","earliestPossibleYear":2023},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"rosa-2018-hlai","url":"https://whocriedagi.com/claims/rosa-2018-hlai","name":"Marek Rosa","role":"CEO and CTO","company":"GoodAI","date":"2018-08-01","datePrecision":"month","quote":"AGI could happen any day.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"At HLAI 2018, he declined a firm deadline.","source":"https://www.weforum.org/stories/emerging-technologies/when-will-we-have-artificial-intelligence-as-smart-as-a-human-here-s-what-experts-think/","sourceLabel":"Futurism / World Economic Forum","evidence":"Conference interview; exact day unspecified.","target":"Any day; no deadline","expectedArrival":{"sourceDate":"2018-08-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"Any day communicates openness to immediate arrival but supplies no bounded interval or deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2017-wired","url":"https://whocriedagi.com/claims/kurzweil-2017-wired","name":"Ray Kurzweil","role":"Director of engineering","company":"Google","date":"2017-11-13","datePrecision":"day","quote":"I've been consistent in saying 2029.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Reaffirmed 2029 for a valid Turing test, which he regarded as requiring the full flexibility of human intelligence.","source":"https://www.wired.com/story/ray-kurzweil-on-turing-tests-brain-extenders-and-ai-ethics/","sourceLabel":"WIRED interview","evidence":"Edited transcript of Nicholas Thompson's Council on Foreign Relations interview; publication date used.","target":"2029","expectedArrival":{"sourceDate":"2017-11-13","sourceDatePrecision":"day","kind":"range","startYear":2029,"endYear":2029,"classification":"ahead","interpretation":"The interview reaffirms 2029 as the nominal year for a valid Turing test; it does not provide a probability interval.","estimateType":"named-year"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"bostrom-2017-ieee","url":"https://whocriedagi.com/claims/bostrom-2017-ieee","name":"Nick Bostrom","role":"Director","company":"Future of Humanity Institute","date":"2017-06-01","datePrecision":"month","quote":"Human-level AI within several decades is plausible.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He stressed substantial uncertainty in both directions.","source":"https://spectrum.ieee.org/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away","sourceLabel":"IEEE Spectrum — original interviews","evidence":"June 2017 issue; exact publication day unverified.","target":"A few decades","expectedArrival":{"sourceDate":"2017-06-01","sourceDatePrecision":"month","kind":"unspecified","classification":"unspecified","interpretation":"Several decades is qualitative; no numeric count or calendar window is inferred."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"brooks-2017-ieee","url":"https://whocriedagi.com/claims/brooks-2017-ieee","name":"Rodney Brooks","role":"Chairman and CTO","company":"Rethink Robotics","date":"2017-06-01","datePrecision":"month","quote":"Humanlike intelligence and consciousness remain over a century away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A counterforecast requiring consciousness as well as intelligence.","source":"https://spectrum.ieee.org/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away","sourceLabel":"IEEE Spectrum — original interviews","evidence":"June 2017 issue; exact publication day unverified.","target":"After 2117","expectedArrival":{"sourceDate":"2017-06-01","sourceDatePrecision":"month","kind":"unspecified","classification":"ahead","interpretation":"More than a century from 2017 supplies only a lower marker after 2117, without an upper bound. It is clearly ahead of 2026, but no finite window exists.","earliestYear":2117,"earliestRelation":"after"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2017-ieee","url":"https://whocriedagi.com/claims/kurzweil-2017-ieee","name":"Ray Kurzweil","role":"Co-founder and chancellor","company":"Singularity University","date":"2017-06-01","datePrecision":"month","quote":"Computers should equal and surpass humans' remaining advantages by 2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"An explicit human-capability forecast.","source":"https://spectrum.ieee.org/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away","sourceLabel":"IEEE Spectrum — original interviews","evidence":"June 2017 issue; exact publication day unverified.","target":"2029","expectedArrival":{"sourceDate":"2017-06-01","sourceDatePrecision":"month","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"The prediction explicitly says by 2029."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"marcus-2017-ieee","url":"https://whocriedagi.com/claims/marcus-2017-ieee","name":"Gary Marcus","role":"Professor","company":"New York University","date":"2017-06-01","datePrecision":"month","quote":"Humanlike understanding and reasoning might require another 20–50 years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A longer horizon, specifically for unfamiliar situations.","source":"https://spectrum.ieee.org/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away","sourceLabel":"IEEE Spectrum — original interviews","evidence":"June 2017 issue; exact publication day unverified.","target":"2037–2067","expectedArrival":{"sourceDate":"2017-06-01","sourceDatePrecision":"month","kind":"range","startYear":2037,"endYear":2067,"classification":"ahead","interpretation":"Twenty to fifty years from 2017 gives a long, approximate horizon for unfamiliar-situation reasoning.","approximate":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"rothblatt-2017-ieee","url":"https://whocriedagi.com/claims/rothblatt-2017-ieee","name":"Martine Rothblatt","role":"Founder","company":"Terasem Movement","date":"2017-06-01","datePrecision":"month","quote":"Brain-equivalent computers should arrive this century.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A broad horizon without a narrower deadline.","source":"https://spectrum.ieee.org/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away","sourceLabel":"IEEE Spectrum — original interviews","evidence":"June 2017 issue; exact publication day unverified.","target":"21st century","expectedArrival":{"sourceDate":"2017-06-01","sourceDatePrecision":"month","kind":"by","deadlineYear":2100,"classification":"window-open","interpretation":"This century gives the end of the twenty-first century as a broad upper horizon; it is not a 2001–2100 arrival range."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"brockman-2016-senate","url":"https://whocriedagi.com/claims/brockman-2016-senate","name":"Greg Brockman","role":"Co-founder and CTO","company":"OpenAI","date":"2016-11-30","datePrecision":"day","quote":"General AI could arrive in 10–100 years, with 90% confidence.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A broad uncertainty interval, not an imminent-arrival claim. Brockman said the technology could not be built with existing capabilities.","source":"https://www.govinfo.gov/content/pkg/CHRG-114shrg24175/pdf/CHRG-114shrg24175.pdf","sourceLabel":"U.S. Senate hearing","evidence":"Original hearing transcript, printed page 38 (PDF page 42); November 30, 2016.","target":"2026–2116 (90% interval)","expectedArrival":{"sourceDate":"2016-11-30","sourceDatePrecision":"day","kind":"range","startYear":2026,"endYear":2116,"classification":"unspecified","interpretation":"The 90% interval spans ten to one hundred years after November 2016. Its nominal lower edge is late 2026, so year-only bounds do not establish overlap on October 7.","confidenceInterval":0.9,"approximate":true,"currentYearOverlap":true,"ambiguity":"Current date is close to the relative lower boundary."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"schmidhuber-2016-wired","url":"https://whocriedagi.com/claims/schmidhuber-2016-wired","name":"Jürgen Schmidhuber","role":"Scientific director; co-founder","company":"IDSIA / NNAISENSE","date":"2016-11-23","datePrecision":"day","quote":"Once we have animal-level AI, a few years or decades later we may have human-level AI","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"His own essay forecast animal-level neural AI within a few years, with human-level intelligence potentially following years or decades later.","source":"https://www.wired.com/story/smarter-software-evolve-ai/","sourceLabel":"WIRED — author essay","evidence":"Bylined essay, November 23, 2016. Conditional progression rather than a fixed AGI deadline.","target":"Years or decades after animal-level AI","expectedArrival":{"sourceDate":"2016-11-23","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"The human-level horizon is relative to an undated animal-level milestone. Neither stage supplies enough precision for absolute years."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"hassabis-2015-decades-from-human-level","url":"https://whocriedagi.com/claims/hassabis-2015-decades-from-human-level","name":"Demis Hassabis","role":"Co-founder and CEO","company":"DeepMind","date":"2015-06-22","datePrecision":"day","quote":"We're decades away from anything that's nearing human-level general intelligence.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He separated useful applications expected within five to ten years from human-level general intelligence, which might require numerous further breakthroughs.","source":"https://www.wired.com/story/deepmind/","sourceLabel":"WIRED UK interview with David Rowan","evidence":"Primary reported interview published online June 22, 2015, for the July 2015 issue.","target":"Decades away","expectedArrival":{"sourceDate":"2015-06-22","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Decades away does not supply a numerical interval. The separate five-to-ten-year application forecast is not an AGI date."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"altman-2015-superhuman-decades-away","url":"https://whocriedagi.com/claims/altman-2015-superhuman-decades-away","name":"Sam Altman","role":"President","company":"Y Combinator","date":"2015-03-03","datePrecision":"day","quote":"superhuman machine intelligence (SMI) is probably still decades away","paraphrase":false,"kind":"forecast","topic":"asi","models":[],"context":"Altman argued for beginning safety and governance work despite a potentially long wait. His milestone was superhuman machine intelligence; the post did not announce an existing AGI.","source":"https://blog.samaltman.com/machine-intelligence-part-2","sourceLabel":"Sam Altman: Machine intelligence, part 2","evidence":"Primary author blog. A preserved PDF snapshot at lifearchitect.ai/pdf/2015-Sam-Altman-Machine-Intelligence.pdf displays March 3, 2015 at 6:17 AM; the current page omits its date.","target":"Probably decades away","expectedArrival":{"sourceDate":"2015-03-03","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"Probably decades away is qualitative; no number of decades is inferred."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2014-guardian-emotional-intelligence","url":"https://whocriedagi.com/claims/kurzweil-2014-guardian-emotional-intelligence","name":"Ray Kurzweil","role":"Director of engineering","company":"Google","date":"2014-02-22","datePrecision":"day","quote":"Computers will match human abilities by 2029, including emotion.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Includes learning, humor and conversation.","source":"https://www.theguardian.com/technology/2014/feb/22/robots-google-ray-kurzweil-terminator-singularity-artificial-intelligence","sourceLabel":"The Guardian interview","evidence":"Primary interview; publication date.","target":"2029","expectedArrival":{"sourceDate":"2014-02-22","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"The claim explicitly forecasts matching human abilities by 2029."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2013-big-think-human-intelligence","url":"https://whocriedagi.com/claims/kurzweil-2013-big-think-human-intelligence","name":"Ray Kurzweil","role":"Director of engineering","company":"Google","date":"2013-09-19","datePrecision":"day","quote":"By 2029, they will match human intelligence.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He forecast that machines would combine human-like reasoning with far greater memory and reading capacity. Contemporary systems were examples of progress, not declared AGI.","source":"https://bigthink.com/articles/can-the-brains-metaphor-machine-be-copied/","sourceLabel":"Big Think: Can the Brain's Metaphor Machine Be Copied?","evidence":"Primary In Their Own Words studio transcript published September 19, 2013.","target":"2029","expectedArrival":{"sourceDate":"2013-09-19","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"An explicit by-2029 human-intelligence forecast."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2013-wired-human-machine-gap","url":"https://whocriedagi.com/claims/kurzweil-2013-wired-human-machine-gap","name":"Ray Kurzweil","role":"Director of engineering","company":"Google","date":"2013-04-25","datePrecision":"day","quote":"I believe that gap will close by 2029.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The gap included emotional intelligence and understanding humor as well as language. He did not identify an existing Google product as human-level.","source":"https://www.wired.com/2013/04/kurzweil-google-ai/","sourceLabel":"WIRED interview with Steven Levy","evidence":"Primary interview published April 25, 2013.","target":"2029","expectedArrival":{"sourceDate":"2013-04-25","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"An explicit by-2029 prediction for closing the human-machine capability gap."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2013-fortune-artificial-brain","url":"https://whocriedagi.com/claims/kurzweil-2013-fortune-artificial-brain","name":"Ray Kurzweil","role":"Director of engineering","company":"Google","date":"2013-01-09","datePrecision":"day","quote":"On that question I've consistently said 2029.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Asked when a functioning artificial brain could be built.","source":"https://fortune.com/2013/01/09/can-we-reverse-engineer-the-brain/","sourceLabel":"Fortune interview","evidence":"Primary interview; publication date.","target":"2029","expectedArrival":{"sourceDate":"2013-01-09","sourceDatePrecision":"day","kind":"range","startYear":2029,"endYear":2029,"classification":"ahead","interpretation":"His answer identifies 2029 as the nominal year for a functioning artificial brain; no surrounding interval is supplied.","estimateType":"named-year"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2012-pbs-human-levels","url":"https://whocriedagi.com/claims/kurzweil-2012-pbs-human-levels","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2012-07-10","datePrecision":"day","quote":"Artificial intelligence will reach human levels by around 2029.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"The interview explored longer-term combinations of human and machine intelligence. He presented human-level AI as a future milestone.","source":"https://www.pbs.org/newshour/show/as-humans-and-computers-merge-immortality","sourceLabel":"PBS NewsHour: As Humans and Computers Merge… Immortality?","evidence":"Primary interview transcript available in the search index. July 10, 2012 broadcast date also corroborated by a Boston University Journal of Science & Technology Law citation.","target":"Around 2029","expectedArrival":{"sourceDate":"2012-07-10","sourceDatePrecision":"day","kind":"around","targetYear":2029,"classification":"unspecified","interpretation":"Around 2029 provides an approximate marker. No radius is supplied, so no finite overlap window is fabricated.","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"nilsson-2012-eighty-percent-jobs","url":"https://whocriedagi.com/claims/nilsson-2012-eighty-percent-jobs","name":"Nils Nilsson","role":"Professor emeritus of computer science","company":"Stanford University","date":"2012-01-09","datePrecision":"day","quote":"A 50% chance that machines could perform roughly 80% of human jobs by 2050.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Nilsson explicitly substituted an employment-based criterion for a vague human-level intelligence threshold. His corresponding 10% and 90% dates were 2030 and 2100, assuming no catastrophe.","source":"https://www.lesswrong.com/posts/xoxZdRtpyRnXmhher/q-and-a-with-experts-on-risks-from-ai-2","sourceLabel":"Alexander Kruel: Q&A with experts on risks from AI #2","evidence":"Primary email Q&A published January 9, 2012; entry uses Nilsson's response and his own definition.","target":"50% by 2050; 90% by 2100","expectedArrival":{"sourceDate":"2012-01-09","sourceDatePrecision":"day","kind":"by","deadlineYear":2050,"classification":"window-open","interpretation":"The central estimate is 50% by 2050 for roughly 80% of jobs. The 10% and 90% cumulative dates remain separate milestones, not a single closed arrival interval.","probability":{"value":0.5,"scope":"by-deadline","approximate":true},"probabilityMilestones":[{"deadlineYear":2030,"probability":0.1},{"deadlineYear":2050,"probability":0.5},{"deadlineYear":2100,"probability":0.9}],"condition":"No catastrophe"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2011-proto-agi-eight-years","url":"https://whocriedagi.com/claims/legg-2011-proto-agi-eight-years","name":"Shane Legg","role":"Co-founder","company":"DeepMind","date":"2011-12-31","datePrecision":"day","quote":"Human-level AGI still centered on 2028; an impressive proto-AGI might arrive within eight years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The proposed intermediate system would learn basic perception, movement and language, and tackle unfamiliar simple problems. He distinguished it from full human-level AGI and retained major-disaster caveats.","source":"https://www.vetta.org/2011/12/goodbye-2011-hello-2012/","sourceLabel":"Shane Legg: Goodbye 2011, hello 2012!","evidence":"Primary year-end blog dated December 31, 2011; text retrieved through the search index.","target":"Mean 2028; proto-AGI within eight years","expectedArrival":{"sourceDate":"2011-12-31","sourceDatePrecision":"day","kind":"around","targetYear":2028,"classification":"unspecified","interpretation":"2028 is a mean of a distribution, not a deadline. A separate weaker proto-AGI forecast within eight years is preserved below.","estimateType":"distribution-mean","centralEstimates":{"meanYear":2028},"additionalPredictions":[{"kind":"by","deadlineYear":2019,"classification":"window-elapsed","milestone":"Proto-AGI","interpretation":"Within eight years of the December 2011 post; not the full human-level AGI milestone."}]},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"vinge-2011-superhuman-by-2030","url":"https://whocriedagi.com/claims/vinge-2011-superhuman-by-2030","name":"Vernor Vinge","role":"Retired mathematics and computer science professor; author","company":"San Diego State University","date":"2011-08-01","datePrecision":"day","quote":"I'd be surprised if the Technological Singularity hasn't happened by 2030.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He explicitly excluded catastrophes such as global nuclear war. His threshold was the advent of superhuman intelligence, which he believed could arise through more than one technological route.","source":"https://reactormag.com/a-chat-with-vernor-vinge/","sourceLabel":"Tor.com / Reactor: A Chat with Vernor Vinge","evidence":"Primary interview by John Ottinger III, published August 1, 2011; exact interview day unspecified.","target":"By 2030, barring catastrophe","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2011-fifty-percent-2028","url":"https://whocriedagi.com/claims/legg-2011-fifty-percent-2028","name":"Shane Legg","role":"Co-founder","company":"DeepMind","date":"2011-06-17","datePrecision":"day","quote":"A 50% chance of human-level machine intelligence by 2028.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His 10%, 50% and 90% dates were 2018, 2028 and 2050. The estimate assumed no major disaster and continued favorable political and economic conditions.","source":"https://www.lesswrong.com/posts/No5JpRCHzBrWA4jmS/q-and-a-with-shane-legg-on-risks-from-ai","sourceLabel":"Alexander Kruel: Q&A with Shane Legg","evidence":"Primary email interview, published June 17, 2011, giving explicit probability thresholds.","target":"50% by 2028; 90% by 2050","expectedArrival":{"sourceDate":"2011-06-17","sourceDatePrecision":"day","kind":"by","deadlineYear":2028,"classification":"window-open","interpretation":"The 50% cumulative date is 2028. His 10% and 90% dates are preserved as separate probabilities, not treated as certain bounds.","probability":{"value":0.5,"scope":"by-deadline","approximate":true},"probabilityMilestones":[{"deadlineYear":2018,"probability":0.1},{"deadlineYear":2028,"probability":0.5},{"deadlineYear":2050,"probability":0.9}],"condition":"No major disaster; favorable political and economic conditions"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"schmidhuber-2011-self-improvement-decades","url":"https://whocriedagi.com/claims/schmidhuber-2011-self-improvement-decades","name":"Jürgen Schmidhuber","role":"Scientific director","company":"IDSIA","date":"2011-06-15","datePrecision":"day","quote":"High for the next few decades","paraphrase":false,"kind":"forecast","topic":"agi","models":["Gödel machine"],"context":"He was answering how likely a human-level or subhuman AI was to improve itself rapidly into vastly greater intelligence. He linked theoretical work, including Gödel machines; this was not a claim that such a working system already existed.","source":"https://www.lesswrong.com/posts/BEtQALqgXmL9d9SfE/q-and-a-with-juergen-schmidhuber-on-risks-from-ai","sourceLabel":"Alexander Kruel: Q&A with Jürgen Schmidhuber","evidence":"Primary email interview published June 15, 2011. The excerpt answers a conditional self-improvement question rather than a precise AGI deadline.","target":"The next few decades","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2011-decade-detailed-plan","url":"https://whocriedagi.com/claims/goertzel-2011-decade-detailed-plan","name":"Ben Goertzel","role":"CEO and OpenCog project leader","company":"Novamente / OpenCog","date":"2011-06-10","datePrecision":"day","quote":"if you follow this plan for, let's say, a decade, you'd get there","paraphrase":false,"kind":"forecast","topic":"agi","models":["OpenCog"],"context":"Goertzel said he believed he had a detailed plan for building a human-level thinking machine, while acknowledging that he could be wrong. He distinguished this approach from uploading a particular person's brain.","source":"https://phys.org/news/2011-06-dr-ben-goertzel-artificial-intelligence.html","sourceLabel":"Phys.org interview with Stuart Mason Dambrot","evidence":"Primary interview published June 10, 2011, conducted following Goertzel's talk at the Humanity+ Transhumanism Meets Design conference.","target":"About a decade, if the proposed plan works","expectedArrival":{"sourceDate":"2011-06-10","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"About a decade is conditional on following a proposed plan. The record does not establish when that plan began, so 2021 is not manufactured as a deadline.","relativeDurationYears":10,"condition":"Proposed plan is followed and works"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"sutton-2011-roughly-2030","url":"https://whocriedagi.com/claims/sutton-2011-roughly-2030","name":"Richard Sutton","role":"Professor of computer science","company":"University of Alberta","date":"2011-05-09","datePrecision":"day","quote":"Roughly 2030.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Sutton gave this answer when asked when he expected human-level AI. The short report does not specify a probability or detailed definition.","source":"https://www.vetta.org/2011/05/sutton-on-human-level-ai/","sourceLabel":"Shane Legg's report of Sutton's Gatsby Unit talk","evidence":"Contemporaneous eyewitness account dated May 9, 2011; quotes the audience question and Sutton's answer. Search-index text available.","target":"Around 2030","expectedArrival":{"sourceDate":"2011-05-09","sourceDatePrecision":"day","kind":"around","targetYear":2030,"classification":"unspecified","interpretation":"Roughly 2030 is an approximate marker with no stated interval width.","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2010-forecast-maintained","url":"https://whocriedagi.com/claims/legg-2010-forecast-maintained","name":"Shane Legg","role":"Co-founder","company":"DeepMind","date":"2010-12-10","datePrecision":"day","quote":"Kept his human-level AGI forecast at a 2028 mean and 2025 mode.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The annual update corrected the earlier stated 90% interval to approximately 2018–2045. It remained a probabilistic estimate with considerable uncertainty.","source":"https://www.vetta.org/2010/12/goodbye-2010/","sourceLabel":"Shane Legg: Goodbye 2010","evidence":"Primary author blog dated December 10, 2010; text retrieved through the search index.","target":"Mean 2028; mode 2025; 90% interval 2018–2045","expectedArrival":{"sourceDate":"2010-12-10","sourceDatePrecision":"day","kind":"range","startYear":2018,"endYear":2045,"classification":"window-open","interpretation":"The explicitly stated 90% interval contains 2026. Mean 2028 and mode 2025 are different summaries of the same uncertain distribution, not deadlines.","confidenceInterval":0.9,"approximate":true,"centralEstimates":{"meanYear":2028,"modeYear":2025}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2010-ten-year-funded-effort","url":"https://whocriedagi.com/claims/goertzel-2010-ten-year-funded-effort","name":"Ben Goertzel","role":"CEO and OpenCog co-founder","company":"Novamente / OpenCog","date":"2010-01-01","datePrecision":"year","quote":"An adequately funded, concentrated AGI effort might enable a singularity within roughly a decade.","paraphrase":true,"kind":"forecast","topic":"agi","models":["OpenCog"],"context":"Goertzel distinguished a possible accelerated research program from a forecast of what society would actually do. He acknowledged scientific obstacles and the possibility of a much later outcome.","source":"https://www.thekurzweillibrary.com/we-could-get-to-the-singularity-in-ten-years","sourceLabel":"Goertzel essay, republished by the Kurzweil Library","evidence":"The December 26, 2014 republication explicitly identifies the essay as written in 2010; original month and day are unspecified.","target":"About 2020, conditional on dedicated effort","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"legg-2009-probabilistic-agi-forecast","url":"https://whocriedagi.com/claims/legg-2009-probabilistic-agi-forecast","name":"Shane Legg","role":"Postdoctoral researcher","company":"University College London","date":"2009-12-07","datePrecision":"day","quote":"Human-level AGI was most likely around 2025, with an expected date of 2028.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Legg specified a probability distribution rather than a deadline. He distinguished the mode, or most likely year, from the later mean of his skewed distribution.","source":"https://www.vetta.org/2009/12/tick-tock-tick-tock-bing/","sourceLabel":"Shane Legg: Tick, tock, tick, tock… BING!","evidence":"Primary author blog, dated December 7, 2009; text retrieved from the search index because the legacy site does not always open reliably.","target":"Mode 2025; mean 2028","expectedArrival":{"sourceDate":"2009-12-07","sourceDatePrecision":"day","kind":"around","targetYear":2028,"classification":"unspecified","interpretation":"2028 is the distribution's mean and 2025 its mode. These are not deadlines or the endpoints of a 2025–2028 interval; no bounded window is inferred from them.","estimateType":"distribution-mean","centralEstimates":{"meanYear":2028,"modeYear":2025}},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"vinge-2009-norwescon-on-track","url":"https://whocriedagi.com/claims/vinge-2009-norwescon-on-track","name":"Vernor Vinge","role":"Retired mathematics and computer science professor; author","company":"San Diego State University","date":"2009-10-12","datePrecision":"day","quote":"I think we're ‘on track’.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Asked whether his earlier 2005–2030 window remained plausible, he reaffirmed it. His singularity concept included several routes to superhuman intelligence, including AI and human enhancement.","source":"https://www.norwescon.org/archives/norwescon33/vingeinterview.htm","sourceLabel":"Norwescon 33 interview","evidence":"Primary interview explicitly dated October 12, 2009. This entry records the 2009 reaffirmation, not the original 1993 prediction.","target":"By 2030","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2009-beneficial-agi-path","url":"https://whocriedagi.com/claims/goertzel-2009-beneficial-agi-path","name":"Ben Goertzel","role":"CEO","company":"Novamente","date":"2009-10-01","datePrecision":"month","quote":"A decade or less could be enough for a positive singularity if society made it a major priority.","paraphrase":true,"kind":"forecast","topic":"agi","models":["OpenCog"],"context":"The talk proposed a development path from virtual agents to more capable systems. OpenCog was presented as an approach to pursue, not an already human-level system.","source":"https://goertzel.org/Summit09/Summit_09_text.pdf","sourceLabel":"Singularity Summit 2009 talk text","evidence":"Primary author-hosted transcript, A Path to Beneficial Artificial General Intelligence; title page identifies the 2009 New York summit.","target":"Ten years or less with a concerted societal effort","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2009-computer-history-oral-history","url":"https://whocriedagi.com/claims/kurzweil-2009-computer-history-oral-history","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2009-07-13","datePrecision":"day","quote":"By 2029, machines should combine human cognitive flexibility with superior memory and speed.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"He reiterated that both the required hardware and software would exist in 2029. His threshold encompassed the breadth of human intelligence, not only a narrow task.","source":"https://archive.computerhistory.org/resources/access/text/2013/05/102702108-05-01-acc.pdf","sourceLabel":"Computer History Museum oral history","evidence":"Primary interview by Dag Spicer. Cover dates the recording to July 13, 2009; forecast appears on transcript page 7. The 2013 URL folder is not the statement date.","target":"2029","expectedArrival":{"sourceDate":"2009-07-13","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"The forecast specifies the relevant hardware and software existing by 2029."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2008-cnn-2029","url":"https://whocriedagi.com/claims/kurzweil-2008-cnn-2029","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2008-05-30","datePrecision":"day","quote":"By 2029 a machine will be able to pass a Turing test","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"In a discussion of future human-machine integration, he treated indistinguishable conversational intelligence as the relevant threshold.","source":"https://transcripts.cnn.com/show/gb/date/2008-05-30/segment/01","sourceLabel":"CNN: Glenn Beck transcript","evidence":"Primary broadcast transcript; aired May 30, 2008.","target":"2029","expectedArrival":{"sourceDate":"2008-05-30","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"A Turing-test forecast explicitly stated as by 2029."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"goertzel-2006-ten-years-if-we-try","url":"https://whocriedagi.com/claims/goertzel-2006-ten-years-if-we-try","name":"Ben Goertzel","role":"CEO","company":"Novamente","date":"2006-08-01","datePrecision":"month","quote":"A positive singularity could be achievable within ten years if the right people received sufficient support.","paraphrase":true,"kind":"forecast","topic":"agi","models":["Novamente"],"context":"This was an explicitly conditional possibility, not an expected arrival date. Novamente was a proposed route still requiring substantial development; he acknowledged the plan might fail.","source":"https://www.goertzel.org/papers/tenyears.htm","sourceLabel":"Goertzel: TransVision 2006 talk","evidence":"Author's talk text identifies TransVision in Helsinki, August 2006. Text recovered through the search index; this legacy URL sometimes redirects inconsistently.","target":"Within ten years, conditional on focused effort","createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2006-acm-human-equality","url":"https://whocriedagi.com/claims/kurzweil-2006-acm-human-equality","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2006-01-11","datePrecision":"day","quote":"by 2029 computers will be fully equal to humans and thereafter surpass them","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He contrasted existing narrow AI with a future combination of human cognitive abilities and machines' advantages in speed and memory.","source":"https://www.pressandappearances.com/acm-ubiquity-interviews-ray-kurzweil","sourceLabel":"ACM Ubiquity interview, Kurzweil archive","evidence":"Primary interview reproduced in the author's archive, dated January 11, 2006.","target":"2029","expectedArrival":{"sourceDate":"2006-01-11","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"An explicit by-2029 prediction of human-level capability."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2005-fast-company-turing-test","url":"https://whocriedagi.com/claims/kurzweil-2005-fast-company-turing-test","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2005-03-21","datePrecision":"day","quote":"by 2029 computers will pass the Turing test","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"He described this as machine intelligence becoming behaviorally indistinguishable from biological intelligence. This is a renewed forecast in a separate interview.","source":"https://www.fastcompany.com/72668/search-sixth-sense","sourceLabel":"Fast Company: In Search of the Sixth Sense","evidence":"Primary interview by Lucas Conley, published March 21, 2005.","target":"2029","expectedArrival":{"sourceDate":"2005-03-21","sourceDatePrecision":"day","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"An explicit by-2029 Turing-test prediction."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2002-long-now-wager","url":"https://whocriedagi.com/claims/kurzweil-2002-long-now-wager","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2002-04-01","datePrecision":"month","quote":"Bet that a machine would pass the agreed Turing test by the end of 2029.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"His wager with Mitch Kapor specified lengthy text interviews and judging criteria. Passing this particular test was the proposed proxy for human-level intelligence.","source":"https://www.writingsbyraykurzweil.com/a-wager-on-the-turing-test-the-rules-2","sourceLabel":"Long Now wager rules, Kurzweil archive","evidence":"Primary rules document explicitly says Published April 2002; the archive's July 2020 heading is a republication date.","target":"End of 2029","expectedArrival":{"sourceDate":"2002-04-01","sourceDatePrecision":"month","kind":"by","deadlineYear":2029,"classification":"window-open","interpretation":"The wager explicitly uses the end of 2029, with specified testing rules.","deadlineDate":"2029-12-31"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-2000-richness-of-human-thinking","url":"https://whocriedagi.com/claims/kurzweil-2000-richness-of-human-thinking","name":"Ray Kurzweil","role":"Inventor and futurist","company":"Kurzweil Technologies","date":"2000-06-17","datePrecision":"day","quote":"we will achieve machines in a few decades that have the complexity and richness of human thinking","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"Kurzweil forecast machines with human-like behavioral complexity. He discussed consciousness as a philosophical question rather than claiming a current system was conscious.","source":"https://achievement.org/achiever/ray-kurzwell/","sourceLabel":"Academy of Achievement oral history","evidence":"Primary interview transcript explicitly dated June 17, 2000, Scottsdale, Arizona; distinct from the page's updated biography.","target":"A few decades","expectedArrival":{"sourceDate":"2000-06-17","sourceDatePrecision":"day","kind":"unspecified","classification":"unspecified","interpretation":"A few decades provides no numerical interval. It is not automatically translated into 2020, 2030 or another deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"moravec-2000-robot-intelligence-2040","url":"https://whocriedagi.com/claims/moravec-2000-robot-intelligence-2040","name":"Hans Moravec","role":"Robotics researcher","company":"Carnegie Mellon University","date":"2000-01-13","datePrecision":"day","quote":"Robots could reach human-level intelligence around 2040.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"An incremental robotics forecast.","source":"https://www.theguardian.com/technology/2000/jan/13/onlinesupplement","sourceLabel":"The Guardian interview","evidence":"Contemporaneous interview; publication date.","target":"Around 2040","expectedArrival":{"sourceDate":"2000-01-13","sourceDatePrecision":"day","kind":"around","targetYear":2040,"classification":"unspecified","interpretation":"Around 2040 is a nominal robotics milestone without a stated tolerance band.","approximate":true,"nominalYearRelation":"ahead"},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-1999-merging-mind-machine","url":"https://whocriedagi.com/claims/kurzweil-1999-merging-mind-machine","name":"Ray Kurzweil","role":"Founder and inventor","company":"Kurzweil Technologies","date":"1999-09-01","datePrecision":"month","quote":"By 2029, the software needed for intelligence would be largely mastered.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The article forecast machines expressing a full range of human intellectual and physical skills within several decades. Its separate 2019 date referred to computing power, not completed intelligent software.","source":"https://www.thekurzweillibrary.com/the-coming-merging-of-mind-and-machine","sourceLabel":"The Coming Merging of Mind and Machine, Scientific American (1999)","evidence":"Author's archive explicitly identifies original publication September 1, 1999 and a February 2001 republication.","originalTerminology":"intelligence of machines","target":"2029","expectedArrival":{"kind":"by","deadlineYear":2029,"note":"Software-intelligence forecast; do not substitute the article's earlier hardware-capacity date."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-1999-spiritual-machines","url":"https://whocriedagi.com/claims/kurzweil-1999-spiritual-machines","name":"Ray Kurzweil","role":"Founder and inventor","company":"Kurzweil Technologies","date":"1999-01-01","datePrecision":"year","quote":"By 2029, computers would routinely pass apparently valid Turing tests, while debate about the full breadth of their intelligence continued.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The 2029 chapter forecasts autonomous learning and knowledge creation with little human involvement. It also preserves controversy about whether machines match every aspect of human intelligence; its 2019 forecast was only for disputed test reports.","source":"https://aitskadapa.ac.in/e-books/CSE/COMPUTER%20NETWORKS/The%20Age%20of%20Spiritual%20Machines_%20When%20Computers%20Exceed%20Human%20Intelligence%20%28%20PDFDrive%20%29.pdf","sourceLabel":"The Age of Spiritual Machines (1999), chapter 11","evidence":"Primary book, copyright 1999; 2029 chapter and final timeline distinguish 2019 reports from routine 2029 test success.","originalTerminology":"Turing Test; machine intelligence","target":"2029","expectedArrival":{"kind":"by","deadlineYear":2029,"note":"Routine apparently valid test performance, with continuing debate about intelligence across all domains."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"bostrom-1998-case-for-superintelligence","url":"https://whocriedagi.com/claims/bostrom-1998-case-for-superintelligence","name":"Nick Bostrom","role":"Doctoral researcher in philosophy","company":"London School of Economics","date":"1998-01-01","datePrecision":"year","quote":"There was a case for superhuman artificial intelligence within the first third of the twenty-first century.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"Bostrom assembled conditional arguments about hardware and brain-inspired software. In a 2008 postscript he clarified that this was a case for the possibility, not an all-things-considered prediction above 50% probability.","source":"https://nickbostrom.com/papers/how-long-before-superintelligence/","sourceLabel":"How Long Before Superintelligence? (1998)","evidence":"Author-hosted primary paper: copyright 1997, journal publication 1998, revised October 25, 1998; later postscripts are separately dated.","originalTerminology":"superhuman artificial intelligence","target":"By approximately 2033, as a scenario","expectedArrival":{"kind":"by","deadlineYear":2033,"note":"First third of the century. In 2008 the author explicitly assigned less than 50% probability to this deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"moravec-1997-nova","url":"https://whocriedagi.com/claims/moravec-1997-nova","name":"Hans Moravec","role":"Principal research scientist","company":"Carnegie Mellon University","date":"1997-10-01","datePrecision":"month","quote":"Robots would reach human-level capabilities by the middle of the next century, in software as well as processing power.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Moravec described four generations of robots, with early commercial success funding increasingly sophisticated world models and reasoning. The interview also describes fourth-generation machines roughly forty or fifty years away.","source":"https://www.pbs.org/wgbh/nova/robots/moravec.html","sourceLabel":"NOVA interview with Hans Moravec, October 1997","evidence":"Primary interview transcript explicitly dates the conversation October 1997; page footer dates must not replace this.","originalTerminology":"human level","target":"By mid-century, approximately 2050","expectedArrival":{"kind":"around","targetYear":2050,"note":"The opening answer says middle of the next century; a later answer says forty or fifty years from 1997."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"moravec-1995-no-job-better","url":"https://whocriedagi.com/claims/moravec-1995-no-job-better","name":"Hans Moravec","role":"Principal research scientist","company":"Carnegie Mellon University","date":"1995-10-01","datePrecision":"day","quote":"By around 2040, there will be no job that people can do better than robots.","paraphrase":false,"kind":"forecast","topic":"agi","models":[],"context":"An interview forecasting broadly capable robots after successive improvements in learning and reasoning.","source":"https://www.wired.com/1995/10/moravec/","sourceLabel":"Charles Platt interviews Moravec, WIRED (1995)","evidence":"Contemporaneous interview; direct quotation.","originalTerminology":"robots","target":"Around 2040","expectedArrival":{"kind":"around","targetYear":2040,"note":"An approximate date."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"moravec-1993-universal-robot","url":"https://whocriedagi.com/claims/moravec-1993-universal-robot","name":"Hans Moravec","role":"Principal research scientist","company":"Carnegie Mellon University","date":"1993-03-01","datePrecision":"month","quote":"Robots combining world simulation and reasoning could gain most human abilities in the 2030–2040 period.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Moravec proposed generations of robots progressing from simple control to learning, simulation and reasoning. This was a staged research forecast, not a description of an existing universal robot.","source":"https://ntrs.nasa.gov/api/citations/19940022858/downloads/19940022858.pdf","sourceLabel":"The Universal Robot, VISION-21 presentation (1993)","evidence":"Primary conference paper preserved by NASA, pp. 35–41; section Reasoning (2030–2040). Proceedings were published in 1994.","originalTerminology":"universal robot","target":"2030–2040","expectedArrival":{"kind":"range","startYear":2030,"endYear":2040,"note":"The fourth-stage reasoning section predicts most human abilities; a later 2050+ section discusses wider displacement."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"vinge-1993-singularity","url":"https://whocriedagi.com/claims/vinge-1993-singularity","name":"Vernor Vinge","role":"Professor of mathematical sciences","company":"San Diego State University","date":"1993-03-01","datePrecision":"month","quote":"Technology would create greater-than-human intelligence within the next thirty years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The paper also supplied a wider surprise window of 2005–2030. It considered multiple paths, including AI, networks, human-computer interfaces and biological enhancement, so it was not solely an AI-model forecast.","source":"https://frc.ri.cmu.edu/~hpm/book98/com.ch1/vinge.singularity.html","sourceLabel":"The Coming Technological Singularity, VISION-21 (1993)","evidence":"Primary essay identifies the March 30–31, 1993 symposium; section 1 contains both the thirty-year statement and the 2005–2030 window.","originalTerminology":"greater-than-human intelligence","target":"Within thirty years (2023); wider window 2005–2030","expectedArrival":{"kind":"by","deadlineYear":2023,"note":"Thirty years from 1993. The same passage separately says arrival before 2005 or after 2030 would be surprising."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"kurzweil-1990-full-turing-test","url":"https://whocriedagi.com/claims/kurzweil-1990-full-turing-test","name":"Ray Kurzweil","role":"Founder and inventor","company":"Kurzweil Technologies","date":"1990-01-01","datePrecision":"year","quote":"Established, broadly accepted success on a full Turing test would arrive sometime between 2020 and 2070.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Kurzweil distinguished narrow successes and disputed demonstrations from broadly accepted unrestricted intelligent conversation. He explicitly cautioned that his estimate might be inaccurate.","source":"https://www.scribd.com/document/804474225/Ray-Kurzweil-The-Age-of-Intelligent-Machines-The-MIT-Press-1990","sourceLabel":"The Age of Intelligent Machines (1990), pp. 414–415 and timeline","evidence":"Primary book text and final timeline both specify 2020–2070; later retellings sometimes compress this interval.","originalTerminology":"full Turing test","target":"2020–2070","expectedArrival":{"kind":"range","startYear":2020,"endYear":2070,"note":"The book's level 4: broadly accepted full-test success, rather than an initially disputed demonstration."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"waltz-1988-beyond-hardware","url":"https://whocriedagi.com/claims/waltz-1988-beyond-hardware","name":"David Waltz","role":"Professor of computer science and research scientist","company":"Brandeis University / Thinking Machines Corporation","date":"1988-01-01","datePrecision":"year","quote":"Human-level intelligence would require advances extending well beyond the hardware he expected around 2017.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Waltz projected affordable brain-scale hardware around 2017 but warned that software, knowledge and learning could take much longer. Training alone might take twenty years, and multiple such experiments could be necessary; 2037 was not a firm forecast.","source":"https://www.cs.columbia.edu/~waltz/Papers/Prospects%20for%20Truly%20Intelligent%20Machines-Daedalus%201988.pdf","sourceLabel":"The Prospects for Building Truly Intelligent Machines, Daedalus (1988)","evidence":"Author-hosted primary article, Winter 1988, pp. 191–212; hardware/software distinction on pp. 207–208.","originalTerminology":"human-level intelligent machines","target":"Well beyond 2017","expectedArrival":{"kind":"after","startYear":2017,"note":"Open-ended lower-bound indication, not a prediction of AGI in 2017 or a fixed 2037 deadline.","lowerBoundExclusive":true},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"solomonoff-1985-milestones","url":"https://whocriedagi.com/claims/solomonoff-1985-milestones","name":"Ray Solomonoff","role":"AI researcher","company":"Oxbridge Research","date":"1985-06-01","datePrecision":"month","quote":"Human-level problem solving in several technical fields could follow a general AI theory by five to ten years plus a few more.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Solomonoff assigned one to fifty years to a general theory, with two to twenty-five more likely, then additional stages for self-improvement and language. His human-level milestone covered mathematics, science and industrial applications, not every human capability.","source":"https://theworld.com/~rjs/timesc.pdf","sourceLabel":"The Time Scale of Artificial Intelligence (1985)","evidence":"Author-hosted primary paper, Human Systems Management 5, 149–153; milestones B–E. The posted reprint displays a later affiliation.","originalTerminology":"general problem solving capacity near that of a human","target":"After a general theory, another five to ten years plus a few years","expectedArrival":{"kind":"unspecified","note":"A conditional sequence of intervals. The paper does not give a single calendar deadline for its human-level milestone."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"vinge-1983-first-word","url":"https://whocriedagi.com/claims/vinge-1983-first-word","name":"Vernor Vinge","role":"Associate professor of mathematical sciences","company":"San Diego State University","date":"1983-01-01","datePrecision":"month","quote":"Superhuman intelligences should be created between twenty and one hundred years from now.","paraphrase":true,"kind":"forecast","topic":"asi","models":[],"context":"Vinge discussed computer science and biological science, leaving the substrate open. The forecast concerned a transition beyond human intelligence, including possible human-computer augmentation.","source":"https://joshh.ug/195/first_word.html","sourceLabel":"First Word, Omni, January 1983","evidence":"Transcript of Vinge's original Omni essay; the article itself identifies his academic role.","originalTerminology":"superhuman intelligences","target":"2003–2083","expectedArrival":{"kind":"range","startYear":2003,"endYear":2083,"note":"Twenty to one hundred years from January 1983."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"mccarthy-1977-five-to-five-hundred","url":"https://whocriedagi.com/claims/mccarthy-1977-five-to-five-hundred","name":"John McCarthy","role":"Professor of computer science","company":"Stanford University","date":"1977-09-30","datePrecision":"day","quote":"The required breakthroughs could take anywhere from five to five hundred years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"McCarthy emphasized the unpredictability of conceptual breakthroughs. This unusually wide range is a counterpoint to confident near-term forecasts.","source":"https://direct.mit.edu/daed/article/151/2/43/110605/If-We-Succeed","sourceLabel":"1977 McCarthy interview, cited by Stuart Russell","evidence":"Russell quotes the interview and identifies Israel Shenker, Detroit Free Press, September 30, 1977. Original clipping not independently inspected.","originalTerminology":"intelligent machines","target":"Five to five hundred years: 1982–2477","expectedArrival":{"kind":"range","startYear":1982,"endYear":2477,"note":"Very broad interval expressing uncertainty; not an assertion that arrival was imminent."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"moravec-1977-ten-years","url":"https://whocriedagi.com/claims/moravec-1977-ten-years","name":"Hans Moravec","role":"Robotics researcher and doctoral candidate","company":"Stanford University","date":"1977-09-03","datePrecision":"day","quote":"If hardware projections held and software kept pace, machines thinking as well as humans could begin appearing in ten years.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"The essay explicitly made the timing conditional on both hardware and software. It forecast an additional 15 years for human-equivalent computing to become as inexpensive as a pocket calculator.","source":"https://frc.ri.cmu.edu/~hpm/project.archive/general.articles/1977/smart","sourceLabel":"Intelligent Machines: How to Get There from Here (1977)","evidence":"Author-hosted primary essay, dated September 3, 1977; introduction and section 6.","originalTerminology":"human equivalence","target":"Around 1987, conditional","expectedArrival":{"kind":"around","targetYear":1987,"note":"Ten years from the 1977 essay, explicitly conditional on software keeping pace with hardware."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"good-1970-revised-date","url":"https://whocriedagi.com/claims/good-1970-revised-date","name":"I. J. Good","role":"University professor of statistics","company":"Virginia Polytechnic Institute","date":"1970-01-01","datePrecision":"year","quote":"An ultraintelligent machine might arrive around 1994, with ten years of uncertainty in either direction.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"An approximate interval, not a firm deadline.","source":"https://coefficientgiving.org/research/what-should-we-learn-from-past-ai-forecasts/","sourceLabel":"Some Future Social Repercussions of Computers (1970)","evidence":"Secondary review cites the original paper's page 69.","originalTerminology":"ultraintelligent machine","target":"1984–2004","expectedArrival":{"kind":"range","startYear":1984,"endYear":2004,"note":"Original estimate: approximately 1994, plus or minus ten years."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"minsky-1967-generation","url":"https://whocriedagi.com/claims/minsky-1967-generation","name":"Marvin Minsky","role":"Professor of electrical engineering","company":"MIT","date":"1967-01-01","datePrecision":"year","quote":"Within a generation, artificial intelligence would be substantially solved and few intellectual domains would remain exclusively human.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Minsky extrapolated from early game-playing, learning and mathematical programs. He did not define the length of a generation, so no numerical arrival year is inferred.","source":"https://quoteinvestigator.com/2021/03/04/ai-solved/","sourceLabel":"Computation: Finite and Infinite Machines (1967), p. 2","evidence":"Quote Investigator checked the original book scans and reproduces the passage with full bibliographic attribution.","originalTerminology":"artificial intelligence","target":"Within a generation","expectedArrival":{"kind":"unspecified","note":"The source supplies no generation length; assigning 1987 or 1992 would add an assumption."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"good-1965-ultraintelligence","url":"https://whocriedagi.com/claims/good-1965-ultraintelligence","name":"I. J. Good","role":"Senior research fellow","company":"Trinity College, Oxford / Atlas Computer Laboratory","date":"1965-01-01","datePrecision":"year","quote":"It was more likely than not that an ultraintelligent machine would be built within the twentieth century.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Good defined the target as exceeding any human across intellectual activities, including the design of better machines. His proposed intelligence explosion followed that first machine; it was not a claim that contemporary systems had reached this level.","source":"https://scienceplusplus.org/metascience/assets/Good1965.pdf","sourceLabel":"Speculations Concerning the First Ultraintelligent Machine (1965)","evidence":"Primary paper, Advances in Computers 6, pp. 31–88; prediction in the concluding discussion, p. 78.","originalTerminology":"ultraintelligent machine","target":"By the end of the twentieth century","expectedArrival":{"kind":"by","deadlineYear":2000,"note":"The source says within the twentieth century, not a particular month or day."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"good-1962-newton-machine","url":"https://whocriedagi.com/claims/good-1962-newton-machine","name":"I. J. Good","role":"Mathematician and statistician","company":"Admiralty Research Laboratory / Institute for Defense Analyses","date":"1962-01-01","datePrecision":"year","quote":"An exceptional machine intellect capable of redesigning itself might emerge in 1978.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Speculation about educating a machine beyond exceptional human ability.","source":"https://coefficientgiving.org/research/what-should-we-learn-from-past-ai-forecasts/","sourceLabel":"The Scientist Speculates (1962), reproduced in historical research","evidence":"Secondary transcription; IEEE's Good biography independently confirms the book's 1962 publication.","originalTerminology":"baby Newton","target":"1978","expectedArrival":{"kind":"around","targetYear":1978,"note":"Explicit speculative estimate."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"minsky-1961-lifetime","url":"https://whocriedagi.com/claims/minsky-1961-lifetime","name":"Marvin Minsky","role":"Co-founder, AI research group","company":"MIT","date":"1961-01-01","datePrecision":"year","quote":"Machines might exceed human general intelligence within a lifetime.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A possibility statement with no calendar deadline.","source":"https://coefficientgiving.org/research/what-should-we-learn-from-past-ai-forecasts/","sourceLabel":"Minsky's 1961 statement, reproduced in historical research","evidence":"Secondary review reproduces the statement; original publication not independently inspected.","originalTerminology":"general intelligence","target":"Within our lifetime","expectedArrival":{"kind":"unspecified","note":"No numeric deadline supplied."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"simon-1960-any-work","url":"https://whocriedagi.com/claims/simon-1960-any-work","name":"Herbert Simon","role":"Professor of administration","company":"Carnegie Institute of Technology","date":"1960-04-07","datePrecision":"day","quote":"Within twenty years, machines would be technically capable of any work a person could do.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Simon distinguished technological capability from economic adoption. He expected people to retain comparative advantages in flexible manual work, some problem solving and interpersonal services.","source":"https://www.scribd.com/document/696737510/Simon-H-A-1960-The-new-science-of-management-decision-Herbert-A-Simon-z-lib-org","sourceLabel":"The New Science of Management Decision (1960), lecture III","evidence":"Primary book p. 38; lecture date April 7, 1960 independently verified from scans by Quote Investigator. Often misdated to its 1965 republication.","originalTerminology":"machines capable of any human work","target":"Within twenty years: 1980","expectedArrival":{"kind":"by","deadlineYear":1980,"note":"Twenty years from the original 1960 lecture, not twenty years from the 1965 republication."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"good-1959-artificial-brain","url":"https://whocriedagi.com/claims/good-1959-artificial-brain","name":"I. J. Good","role":"Mathematician and statistician","company":"Royal Naval Scientific Service","date":"1959-06-02","datePrecision":"day","quote":"A useful artificial brain could take roughly 13 to 30 years, conditional on it being achieved within the next century.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Good gave a central estimate of 20 years multiplied or divided by 1.5, conditional on development within 100 years, which he considered more likely than not. A sufficiently capable machine could then design a better successor.","source":"https://gwern.net/doc/ai/nn/1959-good.pdf","sourceLabel":"Speculations on Perceptrons and Other Automata, IBM RC-115","evidence":"Primary report dated June 2, 1959, pp. 16–17; based on a December 17, 1958 lecture, whose exact wording was not separately verified.","originalTerminology":"artificial brain","target":"Approximately 1972–1989, conditional","expectedArrival":{"kind":"range","startYear":1972,"endYear":1989,"note":"1959 plus 20/1.5 to 20×1.5 years, rounded to calendar years; central estimate 1979. Conditional on achievement within a century."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"simon-newell-1957-thinking-machines","url":"https://whocriedagi.com/claims/simon-newell-1957-thinking-machines","name":"Herbert Simon","role":"Professor of administration","company":"Carnegie Institute of Technology","date":"1957-11-14","datePrecision":"day","quote":"There are now machines that think, learn and create; their range of problems will expand toward the human range.","paraphrase":true,"kind":"qualified","topic":"agi","models":["Logic Theorist"],"context":"Simon's address presented work co-authored with Allen Newell, including their theorem-proving program. The assertion of existing thought was broader than the demonstrated tasks; the complete human problem range remained a future expectation without a date.","source":"https://cse.buffalo.edu/~rapaport/584/S07/02Feb2007Readings.pdf","sourceLabel":"Simon and Newell, Heuristic Problem Solving (1957 address / 1958 paper)","evidence":"Primary paper; publisher's abstract confirms Simon delivered the joint address November 14, 1957. Publication followed in January–February 1958.","originalTerminology":"machines that think, learn, and create","target":"Human problem range in the foreseeable future","expectedArrival":{"kind":"unspecified","note":"Existing thinking-machine claim was qualified; future generality had no numerical deadline."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"turing-1952-unrestricted-test","url":"https://whocriedagi.com/claims/turing-1952-unrestricted-test","name":"Alan Turing","role":"Reader in mathematics","company":"University of Manchester","date":"1952-01-14","datePrecision":"day","quote":"An unrestricted imitation-game contest would be at least a century away.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"Responding to Max Newman's question about a machine having a chance with unrestricted questions, Turing gave a lower bound, not a deadline. This differs from his limited five-minute test prediction in 1950.","source":"https://spectrum.ieee.org/what-turing-himself-said-about-the-imitation-game","sourceLabel":"BBC 1952 discussion, reproduced and examined by Turing scholars","evidence":"Copeland and Proudfoot reproduce the exchange; broadcast dates January 14 and 23, 1952 are documented in Shieber's The Turing Test.","originalTerminology":"imitation game","target":"2052 or later","expectedArrival":{"kind":"after","startYear":2052,"note":"At least 100 years from the 1952 broadcast; no upper bound.","lowerBoundExclusive":false},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"turing-1951-heretical-theory","url":"https://whocriedagi.com/claims/turing-1951-heretical-theory","name":"Alan Turing","role":"Deputy director, Computing Machine Laboratory","company":"University of Manchester","date":"1951-01-01","datePrecision":"year","quote":"If machines capable of thinking were built, they would probably surpass human intellectual powers soon afterward.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"A conditional forecast of rapid improvement after machine thinking begins. It provides no calendar date for the first thinking machine; the surviving text is dated approximately 1951.","source":"https://gwern.net/doc/ai/1951-turing.pdf","sourceLabel":"Intelligent Machinery, A Heretical Theory (c. 1951)","evidence":"Primary typescript text reprinted in The Essential Turing, pp. 472–475; final paragraph.","originalTerminology":"machine thinking","target":"Soon after machine thinking begins","expectedArrival":{"kind":"unspecified","note":"Conditional sequence, not a dated arrival prediction. Statement date is circa 1951."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"},{"id":"turing-1950-imitation-game","url":"https://whocriedagi.com/claims/turing-1950-imitation-game","name":"Alan Turing","role":"Deputy director, Computing Machine Laboratory","company":"University of Manchester","date":"1950-10-01","datePrecision":"month","quote":"Around 2000, computers could make an average interrogator misidentify them at least 30% of the time in a five-minute imitation game.","paraphrase":true,"kind":"forecast","topic":"agi","models":[],"context":"This was a specific conversational-test prediction with a storage assumption, not a claim that computers would perform every human task. The paper separately expressed hope for competition across intellectual fields.","source":"https://www2.cs.sfu.ca/~vaughan/teaching/415/papers/turing1950.html","sourceLabel":"Computing Machinery and Intelligence, Mind (1950)","evidence":"Primary paper, section 6; original publication Mind 59, October 1950, pp. 433–460.","originalTerminology":"imitation game","target":"About 2000","expectedArrival":{"kind":"around","targetYear":2000,"note":"About fifty years after the 1950 paper; a five-minute conversational benchmark."},"createdAt":"2026-10-07","updatedAt":"2026-10-07"}]}