Doomsday Counter

predictions · evidence · amendments

Doomsday Archive: Predictions, Deadlines & Evidence

The complete collection in one place: failed predictions, AI forecasts, real deadlines, interventions and the far future. Search by a name or source, then narrow the result using only the categories represented here.

A date can describe an expected natural event, a technical limit or a claim that did not occur. The type and assessment travel with each record. How the ledger counts.

35 published records. Search all titles, summaries and source labels below. Dates, scenarios and undated risks keep their own labels.

Date / horizonEntry & sourcesStatus / precision
22 Oct 1844The Great Disappointment of 1844: What Happened?Religious interpretationThe expected return on 22 October 1844 did not occur. The Great Disappointment’s aftermath belongs beside the original Millerite claim.Did not occur as predictedday precision
11–13 Sep 198888 Reasons: the 1988 window and a sequelReligious interpretationEdgar Whisenant’s original rapture window ran from 11–13 September 1988, rather than one exact second. The predicted event did not occur. The 1989 sequel was The Final Shout: Rapture Report 1989, not “89 Reasons”.Did not occur as predictedrange precision
1 Jan 2000Y2K: the deadline that people worked onTechnical deadlineTwo-digit year handling created real risks at the century boundary. Organisations inventoried systems, repaired software and tested it. Limited disruption after preparation is not evidence that the risk was imaginary. Y2K belongs under risk reduced.Risk reducedday precision
10 Sep 2008 · first beamThe LHC: the safety argument has evidenceHistorical contextCERN circulated the first LHC beam on 10 September 2008. This was not the first high-energy collision date. Its safety assessment addressed catastrophe claims using physical arguments, including naturally occurring cosmic-ray collisions. That is a rebuttal story, not an unexplained deadline.Context, not a predictionday precision
21 Oct 2011Camping, October 2011: the linked end dateReligious interpretation21 October 2011 was the physical-end date linked to Camping’s May claim. The predicted end did not occur. It is retained as a second date in the same claim family, with the March 2012 acknowledgement linked, rather than counted as a new author.Did not occur as predictedday precision
21 May 2011Harold Camping’s 2011 Rapture Prediction: What Happened?Religious interpretationA public campaign, a missed rapture and the explanation that followed: Camping’s May and October 2011 predictions in one claim history.Did not occur as predictedday precision
21 Dec 2012December 21, 2012: The Maya Calendar and Apocalypse ClaimsReligious interpretationA Maya calendar cycle boundary fell on 21 December 2012. Modern apocalypse claims attached to it failed; the calendar itself did not.Did not occur as predictedday precision
26 Sep 2022 · demonstrationNASA DART Mission: How the Asteroid Deflection Test WorkedDemonstrated interventionDART struck Dimorphos on 26 September 2022 and changed its motion. Measurements published in 2026 also detected a change in the asteroid pair’s orbit around the Sun. The system was never an Earth-impact threat; this was a planetary-defence test.Event occurredday precision
23–24 Sep 2025Rapture 2025: What Happened to the RaptureTok Prediction?Religious interpretationThe reported RaptureTok claim expected the rapture on 23–24 September 2025. That window passed without the announced event.Did not occur as predictedrange precision
2026 earliest; 2027–2028 later windowessays: 2024 / 2026Dario Amodei’s AI Timeline: The 2026–2028 WindowCapability forecastDario Amodei’s 2024 essay allowed powerful AI as early as 2026; his 2026 risk essay discusses a possible arrival within one or two years. These are conditional capability expectations. “Powerful AI” and human extinction are different outcomes.Awaiting the daterange precision
2027 scenario · forecasts updated separatelyoriginal: Apr 2025AI 2027: a scenario with two endingsScenarioA scenario with race and slowdown endings. The authors’ April and August 2026 capability forecasts are recorded separately; the latest update states its assumptions and does not supply an extinction date.Forecast revisedyear precision
2028 · 50% forecastShane Legg’s 2028 AGI Forecast: What 50% MeansCapability forecastIn a 2011 Q&A, Legg gave 2018, 2028 and 2050 for 10%, 50% and 90% chances of human-level AI, conditional on avoiding major disruption. He reiterated a 50% expectation for 2028 in a 2023 interview. A median is not a promised January day.Awaiting the dateyear precision
13 Apr 2029Apophis 2029: Will It Hit Earth?Scheduled natural eventA 340-metre asteroid will pass about 32,000 km above Earth on 13 April 2029. Better measurements removed the collision concern; spacecraft will investigate what the encounter changes.891 daysUTC calendar days; time unspecified
By 2029 · specified testRay Kurzweil’s 2029 Turing Test Bet: The RulesCapability forecastKurzweil’s wager predicts that a computer will pass a specified Turing test by 2029. The rules, judges and comparison with human participants are part of the claim. Persuasive conversation is not automatically expert performance in every field.Awaiting the dateyear precision
19 Jan 2038 · 03:14:08 UTCYear 2038 Problem: What Happens to Unix Time?Technical deadlineSigned 32-bit Unix timestamps reach their limit on 19 January 2038. Exposure depends on software and data formats; wider representations provide a remedy.4,093 daysExact UTC deadline
2045 · futurist horizonRay Kurzweil’s 2045 Singularity: The Actual ClaimCapability forecastKurzweil’s 2045 singularity is a futurist prediction of profound technological change, generally presented positively by its author. It belongs in a comparison of AI futures, with its definition attached, rather than in a list of promised human extinctions.Awaiting the dateyear precision
Not before 2060Newton’s 2060: a lower bound for a new eraReligious interpretationNewton’s biblical calculations treated 2060 as a possible earliest date for a transformed religious era. They do not establish that Earth will be destroyed in 2060, and they supply no deadline on 1 January.Awaiting the datelower precision
24 Sep 2182Bennu: one date, two different probabilitiesScientific modelNASA’s 2021 calculation gave a 1-in-2,700 impact probability for 24 September 2182 and about 1 in 1,750 cumulatively through 2300. Those refer to different questions. Neither means a collision is scheduled.56,937 daysUTC calendar days; time unspecified
Sep 2240 · local sunset boundaryHebrew year 6000: a calendar and an interpretationReligious interpretationThe beginning of Hebrew year 6001 corresponds to 17 September 2240 in the civil calendar. Jewish days begin at the preceding local sunset. Some readings of Sanhedrin discuss a 6,000-year world; that is not one universal prediction held by all Jews.78,114 daysUTC calendar days; time unspecified
1 Jan 10000 · format-dependentYear 10000: the four-digit calendar has companyTechnical deadlineFour-digit year formats cannot represent year 10000 without an extension. That does not mean every system fails on the same day: formats, calendars and long-lived data differ. RFC 3339’s four-digit syntax is a concrete example.2,912,136 daysUTC calendar days; time unspecified
~1.6–1.86 billion yearsEarth’s plants: the model mattersScientific modelA 2024 climate–weathering study estimates that land plants could persist roughly 1.6–1.86 billion years, depending on feedback assumptions. This revises a simple 600-million-year story. Plant limits, oxygen loss and the end of all life are different questions.Context, not a predictionrange precision
Billions of years · model-dependentAndromeda–Milky Way Collision: Will the Galaxies Merge?Scientific modelA 2025 study found roughly a 50% chance of a Milky Way–Andromeda merger within 10 billion years; a 2026 model found a higher probability and a median around 6.5 billion years. Assumptions matter. The old fixed 4.5-billion-year countdown overstated certainty.Context, not a predictionrange precision
~7.6 billion years · 2008 modelWhen Will the Sun Become a Red Giant? Earth’s FutureScientific modelSchröder and Smith’s 2008 stellar-evolution model places the Sun’s red-giant tip about 7.59 billion years ahead and finds Earth would be engulfed in that model. Habitability is expected to change earlier. This is an approximate model timescale.Context, not a predictionrange precision
2014–2015 eclipse series · no exact doomsday claimAuthor’s Four Blood Moons book description; eclipse series 2014–2015Blood Moon Prophecy: What Did the 2015 Eclipses Mean?Historical contextHagee linked a real eclipse series to prophecy. The checked description gives no exact, testable deadline for the end of the world.Context, not a predictionnone precision
2017 · original website statement not verifiedContemporary TIME report: 25 September 2017David Meade, 2017: reports and missing original wordingHistorical contextMeade’s original statement and exact promised outcome remain unverified in this record.Attribution under reviewnone precision
1910 · return and reported scareHalley in 1910: astronomy and the scareHistorical contextHalley’s 1910 return was a real astronomical event. The earlier entry repeated stories about poison gas, pills and masks without contemporary citations. Those anecdotes need their own sources; a comet’s return is not proof of a specific end-of-world prediction.Context, not a predictionnone precision
No single date · extreme cosmic timescalesHeat Death of the Universe: Is There an End Date?Scientific modelHeat death is a conditional far-future scenario, with no known expiry date. Black-hole evaporation scales do not define one universal countdown.Context, not a predictionnone precision
26 Mar 1997 · discovery dateHeaven’s Gate: people, not a punchlineHistorical contextThe sheriff’s case account records 39 people found dead on 26 March 1997, after deaths over several days. That is a discovery date and a human tragedy, not the due date of a missed calendar prediction. The previous joke has been removed.Context, not a predictionnone precision
2–6 September 1666 · historical eventHistorical event, 2–6 September 1666London, 1666: disaster is not a prophecyHistorical contextThe Great Fire is a documented city disaster. Its dates do not establish a named, advance prediction of the world’s end.Context, not a predictionnone precision
15 May 2003 · date discussed retrospectivelyRetrospective statement dated 23 May 2003Nibiru: The 2003 Prediction and What the Sources ShowHistorical contextA surviving 23 May 2003 text explains missed timing as deliberate deception. Read what it establishes, what it cannot establish, and why later 2012 and 2017 versions stay separate.Context, not a predictionnone precision
July 1999 interpretation · exact day unspecifiedChecked French edition: Le Pelletier, 1867, Centurie X, quatrain LXXIINostradamus 1999 Prediction: What Quatrain X.72 SaysHistorical contextThe French text of X.72 names 1999 and month wording. It supplies no exact day and no clear criterion for Earth’s destruction.Context, not a predictionnone precision
3797 · textual horizon, no catastrophe deadlinePreface dated 1 March 1555; checked translation labelled 1961Nostradamus 3797: Did He Predict the End of the World?Historical contextThe preface to Cesar names 3797 as the reach of Nostradamus’s writing; it does not establish an exact end-of-world deadline.Context, not a predictionnone precision
No exact date · “few thousand days”Sep 2024Sam Altman’s AI Timeline: “A Few Thousand Days”Capability forecastThe Intelligence Age, published in September 2024, says superintelligence may be possible within a few thousand days. It is a possibility claim with a loose horizon. The previous ledger’s 1 January 2034 date was an invented conversion and has been removed.Awaiting the datenone precision
Around the first millennium · reported sermonContemporary account by Abbo of Fleury; no exact sermon date establishedThe year 1000: one sermon is not all EuropeHistorical contextAbbo of Fleury records an apocalyptic sermon and his objection to it. One witness does not establish continent-wide panic.Context, not a predictionnone precision
June 2020 rumour · original post not verifiedContemporary RTVE factcheck: 20 June 2020The “recalculated Maya” rumour of 2020Historical contextThe calendar recalculation is faulty, and contemporary checks did not establish a reliable original post. This is a modern rumour, not a sourced Maya prophecy.Attribution under reviewnone precision

See what changed in the evidence →