OVERHANG

What has been claimed about AI, against what has been measured.

Built 2026-10-19 04:51 UTC · 39 sourced claims · 31 superhuman-ledger events · no LLM anywhere in the pipeline

The Ledger

Every dated prediction about AI: when it was made (dot) and the window it pointed to (bar). The status comes from the data alone. A passed deadline reads "outcome unverified" unless dated evidence resolves it; nothing is marked missed without evidence.

achieved
1
window open
4
open
16
deadline passed: outcome unverified
9
undated
9
said open window open deadline passed, outcome unverified achieved (◆ evidence date)
195019501970197019901990201020102020202020302030204020402050205020602060nowAlan Turing · 1950Alan Turing (1950): By ~2000 a computer will hold an average interrogator to <=70% correct identification after 5 minutes — deadline passed: outcome unverifiedAlan Turing (1950): By ~2000 a computer will hold an average interrogator to <=70% correct identification after 5 minutes — deadline passed: outcome unverifiedAlan Turing · 1950Alan Turing (1950): By ~2000 educated opinion will accept the phrase 'machines thinking' — deadline passed: outcome unverifiedAlan Turing (1950): By ~2000 educated opinion will accept the phrase 'machines thinking' — deadline passed: outcome unverifiedJohn McCarthy et al. · 1955John McCarthy et al. (1955-08-31): A carefully chosen group working together for a summer could make significant advances on machine intelligence problems — undatedHerbert A. Simon and Allen Ne… · 1958Herbert A. Simon and Allen Newell (1958): A computer will be world chess champion within ten years — achievedHerbert A. Simon and Allen Newell (1958): A computer will be world chess champion within ten years — achievedDefeated reigning world chess champion in a standard-time match (1997-05-01)Herbert A. Simon and Allen Ne… · 1958Herbert A. Simon and Allen Newell (1958): A computer will discover and prove an important new theorem within ten years — deadline passed: outcome unverifiedHerbert A. Simon and Allen Newell (1958): A computer will discover and prove an important new theorem within ten years — deadline passed: outcome unverifiedHerbert A. Simon · 1965Herbert A. Simon (1965): Machines will do any work a man can do within twenty years — deadline passed: outcome unverifiedHerbert A. Simon (1965): Machines will do any work a man can do within twenty years — deadline passed: outcome unverifiedI. J. Good · 1965I. J. Good (1965): The first ultraintelligent machine would trigger an intelligence explosion and be humanity's last necessary invention — undatedMarvin Minsky · 1967Marvin Minsky (1967): AI will be substantially solved within a generation — undatedMarvin Minsky (as reported by… · 1970Marvin Minsky (as reported by Life magazine) (1970): General-intelligence machine within 3-8 years — deadline passed: outcome unverifiedMarvin Minsky (as reported by Life magazine) (1970): General-intelligence machine within 3-8 years — deadline passed: outcome unverifiedSir James Lighthill · 1973Sir James Lighthill (1973): Retrospective: AI had not delivered the promised impact (the Lighthill report preceding the first AI winter) — undatedVernor Vinge · 1993Vernor Vinge (1993-03): Technological means for superhuman intelligence within 30 years — deadline passed: outcome unverifiedVernor Vinge (1993-03): Technological means for superhuman intelligence within 30 years — deadline passed: outcome unverifiedVernor Vinge · 1993Vernor Vinge (1993-03): The singularity will occur between 2005 and 2030 — window openVernor Vinge (1993-03): The singularity will occur between 2005 and 2030 — window openHans Moravec · 1998Hans Moravec (1998): Hardware matching the human brain available in cheap machines by the 2020s — window openHans Moravec (1998): Hardware matching the human brain available in cheap machines by the 2020s — window openRay Kurzweil (Long Bets wager… · 2002Ray Kurzweil (Long Bets wager vs Mitchell Kapor) (2002): A machine will pass the Turing test by 2029 — openRay Kurzweil (Long Bets wager vs Mitchell Kapor) (2002): A machine will pass the Turing test by 2029 — openVincent C. Muller and Nick Bo… · 2014Vincent C. Muller and Nick Bostrom (2014): Expert survey (2012/13): 50% chance of HLMI around 2040-2050; 90% by 2075 — openVincent C. Muller and Nick Bostrom (2014): Expert survey (2012/13): 50% chance of HLMI around 2040-2050; 90% by 2075 — openKatja Grace et al. · 2017Katja Grace et al. (2017-05): Expert survey: 50% chance AI outperforms humans in all tasks in 45 years; all jobs automated in 120 years — openKatja Grace et al. (2017-05): Expert survey: 50% chance AI outperforms humans in all tasks in 45 years; all jobs automated in 120 years — openZach Stein-Perlman, Katja Gra… · 2022Zach Stein-Perlman, Katja Grace et al. (2022-08): 2022 survey: 50% chance of HLMI by 2059 — openZach Stein-Perlman, Katja Grace et al. (2022-08): 2022 survey: 50% chance of HLMI by 2059 — openGeoffrey Hinton · 2023Geoffrey Hinton (2023-10-08): AI may reason better than humans in five years — openGeoffrey Hinton (2023-10-08): AI may reason better than humans in five years — openKatja Grace et al. · 2024Katja Grace et al. (2024-01): 2023 survey: 10% by 2027, 50% by 2047 for machines outperforming humans on every task — openKatja Grace et al. (2024-01): 2023 survey: 10% by 2027, 50% by 2047 for machines outperforming humans on every task — openJensen Huang · 2024Jensen Huang (2024-03): AI performs well on every test within five years — openJensen Huang (2024-03): AI performs well on every test within five years — openYann LeCun · 2024Yann LeCun (2024-03-07): Skeptic: generative/LLM approach will not reach human-level AI — undatedElon Musk · 2024Elon Musk (2024-04-09): AI smarter than any one human around end of 2025 — deadline passed: outcome unverifiedElon Musk (2024-04-09): AI smarter than any one human around end of 2025 — deadline passed: outcome unverifiedLeopold Aschenbrenner · 2024Leopold Aschenbrenner (2024-06): AGI by 2027 is strikingly plausible — openLeopold Aschenbrenner (2024-06): AGI by 2027 is strikingly plausible — openSam Altman · 2024Sam Altman (2024-09-23): Superintelligence in 'a few thousand days' — undatedDario Amodei · 2024Dario Amodei (2024-10): Powerful AI could arrive as early as 2026 — window openDario Amodei (2024-10): Powerful AI could arrive as early as 2026 — window openDario Amodei · 2024Dario Amodei (2024-10): Powerful AI compresses 50-100 years of biology progress into 5-10 years — undatedSam Altman · 2025Sam Altman (2025): OpenAI knows how to build AGI; aiming beyond it to superintelligence — undatedSam Altman · 2025Sam Altman (2025): By 2035 one person can command the intellectual capacity of all of 2025 — openSam Altman (2025): By 2035 one person can command the intellectual capacity of all of 2025 — openSam Altman · 2025Sam Altman (2025): Novel-insight systems in 2026 — window openSam Altman (2025): Novel-insight systems in 2026 — window openAnthropic · 2025Anthropic (2025-03-06): Powerful AI systems (Nobel-level across most disciplines) by late 2026 / early 2027 — openAnthropic (2025-03-06): Powerful AI systems (Nobel-level across most disciplines) by late 2026 / early 2027 — openDario Amodei · 2025Dario Amodei (2025-03-10): AI writes 90% of code within 3-6 months — deadline passed: outcome unverifiedDario Amodei (2025-03-10): AI writes 90% of code within 3-6 months — deadline passed: outcome unverifiedDaniel Kokotajlo et al. · 2025Daniel Kokotajlo et al. (2025-04-03): AI 2027 scenario: superhuman AI impact over the next decade exceeds the Industrial Revolution — openDaniel Kokotajlo et al. (2025-04-03): AI 2027 scenario: superhuman AI impact over the next decade exceeds the Industrial Revolution — openDemis Hassabis · 2025Demis Hassabis (2025-04-20): AGI within five to ten years — openDemis Hassabis (2025-04-20): AGI within five to ten years — openDemis Hassabis · 2025Demis Hassabis (2025-04-20): AI could end disease within about a decade — openDemis Hassabis (2025-04-20): AI could end disease within about a decade — openMark Zuckerberg · 2025Mark Zuckerberg (2025-07-30): Superintelligence development is 'now in sight' — undatedMark Zuckerberg · 2025Mark Zuckerberg (2025-07-30): The rest of this decade is decisive for superintelligence's path — openMark Zuckerberg (2025-07-30): The rest of this decade is decisive for superintelligence's path — openSam Altman · 2025Sam Altman (2025-10-28): OpenAI target: intern-level AI research assistant by Sept 2026 — deadline passed: outcome unverifiedSam Altman (2025-10-28): OpenAI target: intern-level AI research assistant by Sept 2026 — deadline passed: outcome unverifiedSam Altman · 2025Sam Altman (2025-10-28): OpenAI target: fully automated AI researcher by 2028 — openSam Altman (2025-10-28): OpenAI target: fully automated AI researcher by 2028 — openDario Amodei · 2026Dario Amodei (2026-01): Powerful AI could be 1-2 years away — openDario Amodei (2026-01): Powerful AI could be 1-2 years away — open

The time axis changes scale at 2015 (dotted line): 1945–2015 is compressed so the crowded recent claims stay legible. Undated claims show the dot only. Hover for the claim; click a name for the quote and source.

Measured

What the claims are measured against. Unbounded series (task length, training compute) get a log-linear fit over frontier points with the window stated; bounded benchmark scores get their best result to date.

How long a task can an AI agent finish? (METR, 50% time horizon)

Frontier now
17 h
claude_mythos_preview_early_inspect · 2026-04-07
Doubling time (our fit)
129 days
n=14 frontier points, 2023-03-14 → 2026-02-05, R²=0.93
METR's published
129 days
METR fit, 2023 onward
If this continues: 1 work-week (40 h)
2026-12
fitted line reaches 1 work-week (40 h)
If this continues: 1 work-month (167 h)
2027-09
fitted line reaches 1 work-month (167 h)
202020222024202620281 s6 s1 min10 min2 h17 h4 wk42 wk1 work-day (8 h)1 work-week (40 h)1 work-month (167 h)nowgpt2 · 2019-02-14 · 3 sdavinci_002 · 2020-05-28 · 9 sgpt_3_5_turbo_instruct · 2022-03-15 · 36 sgpt_4 · 2023-03-14 · 4 mingpt_4_1106_inspect · 2023-11-06 · 4 minclaude_3_opus_inspect · 2024-03-04 · 4 mingpt_4_turbo_inspect · 2024-04-09 · 4 mingpt_4o_inspect · 2024-05-13 · 7 minclaude_3_5_sonnet_20240620_inspect · 2024-06-20 · 11 mino1_preview · 2024-09-12 · 20 minclaude_3_5_sonnet_20241022_inspect · 2024-10-22 · 21 mino1_inspect · 2024-12-05 · 39 minclaude_3_7_sonnet_inspect · 2025-02-24 · 1 ho3_inspect · 2025-04-16 · 2 hclaude_4_opus_inspect · 2025-05-22 · 2 hclaude_4_1_opus_inspect · 2025-08-05 · 2 hgpt_5_2025_08_07_inspect · 2025-08-07 · 3 hgemini_3_pro · 2025-11-18 · 4 hgpt_5_1_codex_max_inspect · 2025-11-19 · 4 hclaude_opus_4_5_inspect · 2025-11-24 · 5 hgpt_5_2 · 2025-12-11 · 6 hclaude_opus_4_6_inspect · 2026-02-05 · 12 hgpt_5_3_codex · 2026-02-05 · 6 hgemini_3_1_pro · 2026-02-19 · 6 hgpt_5_4 · 2026-03-05 · 6 hclaude_mythos_preview_early_inspect · 2026-04-07 · 17 h

Blue: models that set a new record when released. Grey: other models. Solid line: OLS fit of log₂(minutes) on date over the stated window; dashed line: the same fit projected. Following METR, points with a central estimate above 16 hours are left out of the fit; they are still plotted. Source: METR, latest model in file dated 2026-04-07.

Training compute of frontier models (Epoch AI)

Largest reported
9.4e+26 FLOP
GPT-6 Astra · 2026-09-03
Growth (our fit)
4.6× / yr
doubling 166 days; n=16, 2018-05-02 → 2026-09-03
Rows without compute
11 / 138
newest models often undisclosed; not imputed
20102012201420162018202020222024202620281e161e181e201e221e241e26nowKN5 LM + RNN 400/10 (WSJ) · 2010-09-26 · 1e16RNN LM · 2010-09-26 · 1e17RNN 1000/5 + RT09 LM (NIST RT05) · 2010-09-26 · 1e17Deep Autoencoders · 2011-04-29 · 1e17CNN Committee (MNIST) · 2011-09-18 · 1e17Dropout (ImageNet) · 2012-06-03 · 1e17Unsupervised High-level Feature Learner · 2012-07-12 · 1e18AlexNet · 2012-09-30 · 1e18DNN EM segmentation · 2012-12-03 · 1e18DistBelief NNLM · 2013-01-16 · 1e18Visualizing CNNs · 2013-11-12 · 1e18TransE · 2013-12-05 · 1e18SPPNet · 2014-06-18 · 1e19RNNsearch-50* · 2014-09-01 · 1e18VGG16 · 2014-09-04 · 1e19VGG19 · 2014-09-04 · 1e19Seq2Seq LSTM · 2014-09-10 · 1e20SNM-skip · 2014-12-03 · 1e20MSRA (C, PReLU) · 2015-02-06 · 1e19AlphaGo Fan · 2015-10-01 · 1e21SAF R-CNN · 2015-10-28 · 1e19Inception v3 · 2015-12-02 · 1e20DeepSpeech2 (English) · 2015-12-08 · 1e19AlphaGo Lee · 2016-01-27 · 1e21BIG LSTM+CNN INPUTS · 2016-02-11 · 1e20GNMT · 2016-09-26 · 1e22Xception · 2016-10-07 · 1e21NASv3 (CIFAR-10) · 2016-11-05 · 1e21JFT · 2017-07-10 · 1e21OpenAI TI7 DOTA 1v1 · 2017-08-11 · 1e21AlphaGo Zero · 2017-10-18 · 1e21ResNeXt-101 32x48d · 2018-05-02 · 1e22BigGAN-deep 512x512 · 2018-09-28 · 1e21GPT-2 (774M) · 2019-02-14 · 1e22GPT-2 (355M) · 2019-02-14 · 1e21Grover-Mega · 2019-05-29 · 1e22XLNet · 2019-06-01 · 1e22RoBERTa Large · 2019-07-01 · 1e22Megatron-BERT · 2019-09-17 · 1e22Megatron-LM (8.3B) · 2019-09-17 · 1e22T5-3B · 2019-10-23 · 1e22T5-11B · 2019-10-23 · 1e23AlphaStar · 2019-10-30 · 1e23XLM-RoBERTa · 2019-11-05 · 1e22Noisy Student (L2) · 2019-11-11 · 1e22OpenAI Five · 2019-12-13 · 1e23Meena · 2020-01-28 · 1e23GPT-3 175B (davinci) · 2020-05-28 · 1e23iGPT-XL · 2020-06-17 · 1e23GShard (dense) · 2020-06-30 · 1e23mT5-XXL · 2020-10-20 · 1e23Switch · 2021-01-11 · 1e23Wu Dao 2.0 · 2021-05-31 · 1e24Jurassic-1-Jumbo · 2021-08-11 · 1e24FLAN 137B · 2021-09-03 · 1e24HyperCLOVA 204B · 2021-09-10 · 1e23Megatron-Turing NLG 530B · 2021-10-11 · 1e24Yuan 1.0 · 2021-10-12 · 1e24Gopher (280B) · 2021-12-08 · 1e24EXAONE 1.0 · 2021-12-14 · 1e24ERNIE 3.0 Titan · 2021-12-23 · 1e24GPT-3.5 (davinci-002) · 2022-03-15 · 1e24PaLM (540B) · 2022-04-04 · 1e24Minerva (540B) · 2022-06-29 · 1e24U-PaLM (540B) · 2022-10-20 · 1e24Flan-PaLM 540B · 2022-10-20 · 1e24GPT-4 (Mar 2023) · 2023-03-15 · 1e25SenseChat · 2023-04-10 · 1e25PaLM 2 · 2023-05-10 · 1e25GPT-4 (Jun 2023) · 2023-06-13 · 1e25Claude 2 · 2023-07-11 · 1e25Falcon-180B · 2023-09-06 · 1e25Amazon Titan · 2023-09-28 · 1e25Inflection-2 · 2023-11-22 · 1e25Gemini 1.0 Ultra · 2023-12-06 · 1e26Nemotron-4 340B · 2024-06-14 · 1e25Claude 3.5 Sonnet · 2024-06-20 · 1e25Llama 3.1-405B · 2024-07-23 · 1e26Grok-2 · 2024-08-13 · 1e25Grok 3 · 2025-02-17 · 1e27GPT-4.5 · 2025-02-27 · 1e27Llama Nemotron Ultra 253B · 2025-03-18 · 1e26Llama 4 Behemoth (preview) · 2025-04-05 · 1e26Grok 4 · 2025-07-09 · 1e27GPT-6 Astra · 2026-09-03 · 1e27

FLOP, log scale. Fit over frontier (record-setting) points from 2018-01-01. Source: Epoch AI, Frontier AI Models, CC BY 4.0.

Hard benchmarks: best score to date

FrontierMath T1–3
52%
gpt-5.5-pro-pre-release_high · 2026-04-23
ARC-AGI-1
98%
Claude Fable 5 (Max) · 2026-06-09
ARC-AGI-2
95%
GPT-6 Astra (Max) · 2026-09-02
202320242025202620270%25%50%75%100%FrontierMath T1–3 · claude-3-5-sonnet-20240620 · 2024-06-20 · 1%FrontierMath T1–3 · gpt-4o-2024-08-06 · 2024-08-06 · 0%FrontierMath T1–3 · o1-mini-2024-09-12_medium · 2024-09-12 · 2%FrontierMath T1–3 · o1-mini-2024-09-12_high · 2024-09-12 · 1%FrontierMath T1–3 · gemini-1.5-flash-002 · 2024-09-24 · 0%FrontierMath T1–3 · claude-3-5-haiku-20241022 · 2024-10-22 · 0%FrontierMath T1–3 · claude-3-5-sonnet-20241022 · 2024-10-22 · 2%FrontierMath T1–3 · mistral-large-2411 · 2024-11-18 · 0%FrontierMath T1–3 · gpt-4o-2024-11-20 · 2024-11-20 · 0%FrontierMath T1–3 · grok-2-1212 · 2024-12-12 · 1%FrontierMath T1–3 · o1-2024-12-17_high · 2024-12-17 · 9%FrontierMath T1–3 · DeepSeek-V3 · 2024-12-26 · 2%FrontierMath T1–3 · qwen-max-2025-01-25 · 2025-01-25 · 1%FrontierMath T1–3 · o3-mini-2025-01-31_high · 2025-01-31 · 12%FrontierMath T1–3 · o3-mini-2025-01-31_medium · 2025-01-31 · 8%FrontierMath T1–3 · gemini-2.0-flash-001 · 2025-02-05 · 2%FrontierMath T1–3 · claude-3-7-sonnet-20250219_64K · 2025-02-24 · 3%FrontierMath T1–3 · claude-3-7-sonnet-20250219_32K · 2025-02-24 · 3%FrontierMath T1–3 · claude-3-7-sonnet-20250219_16K · 2025-02-24 · 4%FrontierMath T1–3 · claude-3-7-sonnet-20250219 · 2025-02-24 · 3%FrontierMath T1–3 · Llama-4-Maverick-17B-128E-Instruct-FP8 · 2025-04-05 · 1%FrontierMath T1–3 · Llama-4-Scout-17B-16E-Instruct · 2025-04-05 · 0%FrontierMath T1–3 · grok-3-mini-beta_low · 2025-04-09 · 3%FrontierMath T1–3 · grok-3-mini-beta_high · 2025-04-09 · 6%FrontierMath T1–3 · grok-3-beta · 2025-04-09 · 4%FrontierMath T1–3 · gpt-4.1-nano-2025-04-14 · 2025-04-14 · 1%FrontierMath T1–3 · gpt-4.1-mini-2025-04-14 · 2025-04-14 · 4%FrontierMath T1–3 · gpt-4.1-2025-04-14 · 2025-04-14 · 6%FrontierMath T1–3 · o3-2025-04-16_low · 2025-04-16 · 10%FrontierMath T1–3 · o3-2025-04-16_high · 2025-04-16 · 19%FrontierMath T1–3 · o3-2025-04-16_medium · 2025-04-16 · 17%FrontierMath T1–3 · o4-mini-2025-04-16_low · 2025-04-16 · 11%FrontierMath T1–3 · o4-mini-2025-04-16_medium · 2025-04-16 · 19%FrontierMath T1–3 · o4-mini-2025-04-16_high · 2025-04-16 · 25%FrontierMath T1–3 · qwen-plus-2025-04-28 · 2025-04-28 · 2%FrontierMath T1–3 · mistral-medium-2505 · 2025-05-07 · 0%FrontierMath T1–3 · claude-opus-4-20250514_27K · 2025-05-22 · 4%FrontierMath T1–3 · claude-opus-4-20250514 · 2025-05-22 · 4%FrontierMath T1–3 · claude-sonnet-4-20250514 · 2025-05-22 · 4%FrontierMath T1–3 · gemini-2.5-pro-preview-06-05 · 2025-06-05 · 10%FrontierMath T1–3 · gemini-2.5-flash · 2025-06-17 · 5%FrontierMath T1–3 · gemini-2.5-pro · 2025-06-17 · 14%FrontierMath T1–3 · grok-4-0709 · 2025-07-09 · 20%FrontierMath T1–3 · Qwen/Qwen3-235B-A22B-Thinking-2507 · 2025-07-25 · 8%FrontierMath T1–3 · gemini-2.5-deep-think-2025-08-01-webapp · 2025-08-01 · 29%FrontierMath T1–3 · claude-opus-4-1-20250805_27K · 2025-08-05 · 7%FrontierMath T1–3 · claude-opus-4-1-20250805 · 2025-08-05 · 6%FrontierMath T1–3 · gpt-5-2025-08-07_high · 2025-08-07 · 32%FrontierMath T1–3 · gpt-5-mini-2025-08-07_high · 2025-08-07 · 27%FrontierMath T1–3 · gpt-5-mini-2025-08-07_medium · 2025-08-07 · 20%FrontierMath T1–3 · gpt-5-2025-08-07_medium · 2025-08-07 · 27%FrontierMath T1–3 · gpt-5-nano-2025-08-07_high · 2025-08-07 · 8%FrontierMath T1–3 · gpt-5-nano-2025-08-07_medium · 2025-08-07 · 7%FrontierMath T1–3 · claude-sonnet-4-5-20250929 · 2025-09-29 · 9%FrontierMath T1–3 · claude-sonnet-4-5-20250929_32K · 2025-09-29 · 15%FrontierMath T1–3 · claude-sonnet-4-5-20250929_59K · 2025-09-29 · 13%FrontierMath T1–3 · zai-org/GLM-4.6 · 2025-09-30 · 4%FrontierMath T1–3 · claude-haiku-4-5-20251001_32K · 2025-10-15 · 6%FrontierMath T1–3 · claude-haiku-4-5-20251001 · 2025-10-15 · 4%FrontierMath T1–3 · moonshotai/Kimi-K2-Thinking · 2025-11-06 · 21%FrontierMath T1–3 · gpt-5.1-2025-11-13_none · 2025-11-13 · 2%FrontierMath T1–3 · gpt-5.1-2025-11-13_low · 2025-11-13 · 17%FrontierMath T1–3 · gpt-5.1-2025-11-13_medium · 2025-11-13 · 27%FrontierMath T1–3 · gpt-5.1-2025-11-13_high · 2025-11-13 · 31%FrontierMath T1–3 · gemini-3-pro-preview · 2025-11-18 · 38%FrontierMath T1–3 · claude-opus-4-5-20251101 · 2025-11-24 · 21%FrontierMath T1–3 · claude-opus-4-5-20251101_32K · 2025-11-24 · 21%FrontierMath T1–3 · claude-opus-4-5-20251101_16K · 2025-11-24 · 20%FrontierMath T1–3 · fireworks/deepseek-v3p2 · 2025-12-01 · 22%FrontierMath T1–3 · gpt-5.2-2025-12-11_xhigh · 2025-12-11 · 41%FrontierMath T1–3 · gpt-5.2-2025-12-11_high · 2025-12-11 · 40%FrontierMath T1–3 · gpt-5.2-2025-12-11_medium · 2025-12-11 · 37%FrontierMath T1–3 · gpt-5.2-2025-12-11_low · 2025-12-11 · 27%FrontierMath T1–3 · gemini-3-flash-preview · 2025-12-17 · 36%FrontierMath T1–3 · zai-org/GLM-4.7 · 2025-12-22 · 2%FrontierMath T1–3 · fireworks/kimi-k2p5 · 2026-01-27 · 28%FrontierMath T1–3 · claude-opus-4-6_max · 2026-02-05 · 41%FrontierMath T1–3 · claude-opus-4-6_64K · 2026-02-05 · 40%FrontierMath T1–3 · claude-opus-4-6_32K · 2026-02-05 · 40%FrontierMath T1–3 · claude-opus-4-6 · 2026-02-05 · 38%FrontierMath T1–3 · glm-5 · 2026-02-11 · 16%FrontierMath T1–3 · qwen3.5-plus · 2026-02-16 · 21%FrontierMath T1–3 · claude-sonnet-4-6_16K · 2026-02-17 · 32%FrontierMath T1–3 · gemini-3.1-pro-preview · 2026-02-19 · 37%FrontierMath T1–3 · qwen3.5-flash · 2026-02-25 · 6%FrontierMath T1–3 · gpt-5.4-2026-03-05_xhigh · 2026-03-05 · 48%FrontierMath T1–3 · gpt-5.4-pro-2026-03-05_xhigh · 2026-03-05 · 50%FrontierMath T1–3 · gpt-5.4-mini-2026-03-17_high · 2026-03-17 · 28%FrontierMath T1–3 · gpt-5.4-nano-2026-03-17_high · 2026-03-17 · 26%FrontierMath T1–3 · qwen3.6-plus · 2026-03-31 · 26%FrontierMath T1–3 · glm-5.1 · 2026-04-07 · 33%FrontierMath T1–3 · muse-spark · 2026-04-08 · 39%FrontierMath T1–3 · claude-opus-4-7_xhigh · 2026-04-16 · 44%FrontierMath T1–3 · qwen3.6-max-preview · 2026-04-20 · 23%FrontierMath T1–3 · kimi-k2.6 · 2026-04-20 · 39%FrontierMath T1–3 · gpt-5.5-pro-pre-release_xhigh · 2026-04-23 · 51%FrontierMath T1–3 · gpt-5.5-pro-pre-release_high · 2026-04-23 · 52%FrontierMath T1–3 · gpt-5.5-pre-release_xhigh · 2026-04-23 · 52%FrontierMath T1–3 · qwen3.6-flash · 2026-04-27 · 10%FrontierMath T1–3 · gemini-3.5-flash_high · 2026-05-19 · 39%FrontierMath T1–3 · claude-opus-4-8_max · 2026-05-28 · 47%FrontierMath T1–3ARC-AGI-2: Human Panel 100%ARC-AGI-2 · Icecuber · 2023-11-03 · 2%ARC-AGI-2 · GPT-4o-mini · 2024-07-18 · 0%ARC-AGI-2 · o1-mini · 2024-09-12 · 1%ARC-AGI-2 · NVARC · 2024-11-03 · 28%ARC-AGI-2 · ARChitects · 2024-11-03 · 2%ARC-AGI-2 · GPT-4o · 2024-11-20 · 0%ARC-AGI-2 · o3 (Preview, Low) ¹ · 2024-12-20 · 4%ARC-AGI-2 · Deepseek R1 · 2025-01-22 · 1%ARC-AGI-2 · o3-mini (Low) · 2025-01-31 · 0%ARC-AGI-2 · o3-mini (Medium) · 2025-01-31 · 2%ARC-AGI-2 · o3-mini (High) · 2025-01-31 · 3%ARC-AGI-2 · Gemini 2.0 Flash · 2025-02-02 · 1%ARC-AGI-2 · Gemini 1.5 Pro · 2025-02-15 · 1%ARC-AGI-2 · Grok 3 · 2025-02-19 · 0%ARC-AGI-2 · Grok 3 Mini (Low) · 2025-02-19 · 0%ARC-AGI-2 · Gemini 2.5 Pro (Preview) · 2025-02-19 · 4%ARC-AGI-2 · Gemini 2.5 Pro (Preview, Thinking 1K) · 2025-02-19 · 3%ARC-AGI-2 · Claude 3.7 · 2025-02-24 · 0%ARC-AGI-2 · Claude 3.7 (16K) · 2025-02-24 · 1%ARC-AGI-2 · Claude 3.7 (1K) · 2025-02-24 · 0%ARC-AGI-2 · Claude 3.7 (8K) · 2025-02-24 · 1%ARC-AGI-2 · GPT-4.5 · 2025-02-27 · 1%ARC-AGI-2 · Llama 4 Maverick · 2025-04-05 · 0%ARC-AGI-2 · Llama 4 Scout · 2025-04-05 · 0%ARC-AGI-2 · GPT-4.1-Nano · 2025-04-14 · 0%ARC-AGI-2 · GPT-4.1-Mini · 2025-04-14 · 0%ARC-AGI-2 · GPT-4.1 · 2025-04-14 · 0%ARC-AGI-2 · o3 (Low) · 2025-04-16 · 2%ARC-AGI-2 · o3 (Medium) · 2025-04-16 · 3%ARC-AGI-2 · o3 (High) · 2025-04-16 · 7%ARC-AGI-2 · o4-mini (Low) · 2025-04-16 · 2%ARC-AGI-2 · o4-mini (Medium) · 2025-04-16 · 2%ARC-AGI-2 · o4-mini (High) · 2025-04-16 · 6%ARC-AGI-2 · Claude Sonnet 4 · 2025-05-14 · 1%ARC-AGI-2 · Claude Sonnet 4 (Thinking 1K) · 2025-05-14 · 1%ARC-AGI-2 · Claude Sonnet 4 (Thinking 8K) · 2025-05-14 · 2%ARC-AGI-2 · Claude Sonnet 4 (Thinking 16K) · 2025-05-14 · 6%ARC-AGI-2 · Claude Opus 4 (Thinking 16K) · 2025-05-14 · 9%ARC-AGI-2 · Claude Opus 4 (Thinking 8K) · 2025-05-14 · 5%ARC-AGI-2 · Claude Opus 4 (Thinking 1K) · 2025-05-14 · 0%ARC-AGI-2 · Claude Opus 4 · 2025-05-14 · 1%ARC-AGI-2 · Gemini 2.5 Flash (Preview) · 2025-05-20 · 2%ARC-AGI-2 · Gemini 2.5 Flash (Preview) (Thinking 1K) · 2025-05-20 · 2%ARC-AGI-2 · Gemini 2.5 Flash (Preview) (Thinking 8K) · 2025-05-20 · 2%ARC-AGI-2 · Gemini 2.5 Flash (Preview) (Thinking 16K) · 2025-05-20 · 2%ARC-AGI-2 · Gemini 2.5 Flash (Preview) (Thinking 24K) · 2025-05-20 · 3%ARC-AGI-2 · Codex Mini (Latest) · 2025-05-23 · 1%ARC-AGI-2 · Deepseek R1 (05/28) · 2025-05-28 · 1%ARC-AGI-2 · o3-Pro (Low) · 2025-06-10 · 2%ARC-AGI-2 · o3-Pro (Medium) · 2025-06-10 · 2%ARC-AGI-2 · o3-Pro (High) · 2025-06-10 · 5%ARC-AGI-2 · Magistral Small · 2025-06-10 · 0%ARC-AGI-2 · Magistral Medium · 2025-06-10 · 0%ARC-AGI-2 · Magistral Medium (Thinking) · 2025-06-10 · 0%ARC-AGI-2 · Gemini 2.5 Pro (Thinking 1K) · 2025-06-17 · 0%ARC-AGI-2 · Gemini 2.5 Pro (Thinking 8K) · 2025-06-17 · 3%ARC-AGI-2 · Gemini 2.5 Pro (Thinking 16K) · 2025-06-17 · 4%ARC-AGI-2 · Gemini 2.5 Pro (Thinking 32K) · 2025-06-17 · 5%ARC-AGI-2 · Grok 4 (Thinking) · 2025-07-09 · 16%ARC-AGI-2 · Qwen3-235b-a22b Instruct (25/07) · 2025-07-25 · 1%ARC-AGI-2 · GPT-5 (High) · 2025-08-07 · 10%ARC-AGI-2 · GPT-5 (Medium) · 2025-08-07 · 7%ARC-AGI-2 · GPT-5 (Low) · 2025-08-07 · 2%ARC-AGI-2 · GPT-5 (Low) · 2025-08-07 · 2%ARC-AGI-2 · GPT-5 (Minimal) · 2025-08-07 · 0%ARC-AGI-2 · GPT-5 Mini (High) · 2025-08-07 · 4%ARC-AGI-2 · GPT-5 Mini (Medium) · 2025-08-07 · 4%ARC-AGI-2 · GPT-5 Mini (Low) · 2025-08-07 · 1%ARC-AGI-2 · GPT-5 Mini (Minimal) · 2025-08-07 · 2%ARC-AGI-2 · GPT-5 Nano (High) · 2025-08-07 · 3%ARC-AGI-2 · GPT-5 Nano (Medium) · 2025-08-07 · 1%ARC-AGI-2 · GPT-5 Nano (Low) · 2025-08-07 · 0%ARC-AGI-2 · GPT-5 Nano (Minimal) · 2025-08-07 · 0%ARC-AGI-2 · Hierarchical Reasoning Model (HRM) · 2025-08-07 · 2%ARC-AGI-2 · Grok 4 (Refine.) · 2025-08-07 · 29%ARC-AGI-2 · Grok 4 (Refine.) · 2025-08-07 · 26%ARC-AGI-2 · Claude Sonnet 4.5 · 2025-09-29 · 4%ARC-AGI-2 · Claude Sonnet 4.5 (Thinking 1K) · 2025-09-29 · 6%ARC-AGI-2 · Claude Sonnet 4.5 (Thinking 8K) · 2025-09-29 · 7%ARC-AGI-2 · Claude Sonnet 4.5 (Thinking 16K) · 2025-09-29 · 7%ARC-AGI-2 · Claude Sonnet 4.5 (Thinking 32K) · 2025-09-29 · 14%ARC-AGI-2 · Claude Haiku 4.5 · 2025-10-01 · 1%ARC-AGI-2 · Claude Haiku 4.5 (Thinking 1K) · 2025-10-01 · 1%ARC-AGI-2 · Claude Haiku 4.5 (Thinking 8K) · 2025-10-01 · 2%ARC-AGI-2 · Claude Haiku 4.5 (Thinking 16K) · 2025-10-01 · 3%ARC-AGI-2 · Claude Haiku 4.5 (Thinking 32K) · 2025-10-01 · 4%ARC-AGI-2 · GPT-5 Pro · 2025-10-06 · 18%ARC-AGI-2 · Tiny Recursion Model (TRM) · 2025-10-07 · 6%ARC-AGI-2 · Grok 4 (Fast Reasoning) · 2025-10-21 · 5%ARC-AGI-2 · GPT-5.1 (Thinking, None) · 2025-11-13 · 0%ARC-AGI-2 · GPT-5.1 (Thinking, Low) · 2025-11-13 · 2%ARC-AGI-2 · GPT-5.1 (Thinking, Medium) · 2025-11-13 · 7%ARC-AGI-2 · GPT-5.1 (Thinking, High) · 2025-11-13 · 18%ARC-AGI-2 · Gemini 3 Pro · 2025-11-18 · 31%ARC-AGI-2 · Opus 4.5 (Thinking, None) · 2025-11-24 · 8%ARC-AGI-2 · Opus 4.5 (Thinking, 1K) · 2025-11-24 · 9%ARC-AGI-2 · Opus 4.5 (Thinking, 8K) · 2025-11-24 · 14%ARC-AGI-2 · Opus 4.5 (Thinking, 16K) · 2025-11-24 · 23%ARC-AGI-2 · Opus 4.5 (Thinking, 32K) · 2025-11-24 · 31%ARC-AGI-2 · Opus 4.5 (Thinking, 64K) · 2025-11-24 · 38%ARC-AGI-2 · Deepseek V3.2 · 2025-12-01 · 4%ARC-AGI-2 · Gemini 3 Pro (Refine.) · 2025-12-04 · 54%ARC-AGI-2 · GPT-5.2 · 2025-12-11 · 1%ARC-AGI-2 · GPT-5.2 (Low) · 2025-12-11 · 10%ARC-AGI-2 · GPT-5.2 (Medium) · 2025-12-11 · 27%ARC-AGI-2 · GPT-5.2 (High) · 2025-12-11 · 43%ARC-AGI-2 · GPT-5.2 (XHigh) · 2025-12-11 · 53%ARC-AGI-2 · GPT-5.2 Pro (Medium) · 2025-12-11 · 38%ARC-AGI-2 · GPT-5.2 Pro (High) · 2025-12-11 · 54%ARC-AGI-2 · Gemini 3 Flash Preview (Minimal) · 2025-12-17 · 3%ARC-AGI-2 · Gemini 3 Flash Preview (Low) · 2025-12-17 · 1%ARC-AGI-2 · Gemini 3 Flash Preview (Medium) · 2025-12-17 · 13%ARC-AGI-2 · Gemini 3 Flash Preview (High) · 2025-12-17 · 34%ARC-AGI-2 · Kimi K2.5 · 2026-01-27 · 12%ARC-AGI-2 · GPT-5.2 (Refine.) · 2026-02-03 · 73%ARC-AGI-2 · Claude Opus 4.6 (120K, Low) · 2026-02-05 · 65%ARC-AGI-2 · Claude Opus 4.6 (120K, Medium) · 2026-02-05 · 66%ARC-AGI-2 · Claude Opus 4.6 (120K, High) · 2026-02-05 · 69%ARC-AGI-2 · Claude Opus 4.6 (120K, Max) · 2026-02-05 · 69%ARC-AGI-2 · Gemini 3 Deep Think (Preview) ² · 2026-02-12 · 45%ARC-AGI-2 · Gemini 3 Deep Think (2/26) · 2026-02-12 · 85%ARC-AGI-2 · Minimax M2.5 · 2026-02-12 · 5%ARC-AGI-2 · Claude Sonnet 4.6 (Max) · 2026-02-17 · 58%ARC-AGI-2 · Claude Sonnet 4.6 (High) · 2026-02-17 · 60%ARC-AGI-2 · Gemini 3.1 Pro (Preview) · 2026-02-19 · 77%ARC-AGI-2 · GLM-5 · 2026-02-20 · 5%ARC-AGI-2 · GPT-5.4 (Low) · 2026-03-04 · 29%ARC-AGI-2 · GPT-5.4 (Medium) · 2026-03-04 · 55%ARC-AGI-2 · GPT-5.4 (High) · 2026-03-04 · 68%ARC-AGI-2 · GPT-5.4 (XHigh) · 2026-03-04 · 74%ARC-AGI-2 · GPT-5.4 Pro (XHigh) · 2026-03-04 · 83%ARC-AGI-2 · Grok 4.20 (Reasoning) · 2026-03-09 · 65%ARC-AGI-2 · GPT-5.4 Mini (XHigh) · 2026-03-17 · 19%ARC-AGI-2 · GPT-5.4 Mini (High) · 2026-03-17 · 13%ARC-AGI-2 · GPT-5.4 Mini (Medium) · 2026-03-17 · 4%ARC-AGI-2 · GPT-5.4 Mini (Low) · 2026-03-17 · 1%ARC-AGI-2 · GPT-5.4 Nano (XHigh) · 2026-03-17 · 6%ARC-AGI-2 · GPT-5.4 Nano (High) · 2026-03-17 · 4%ARC-AGI-2 · GPT-5.4 Nano (Medium) · 2026-03-17 · 2%ARC-AGI-2 · GPT-5.4 Nano (Low) · 2026-03-17 · 2%ARC-AGI-2 · Claude 4.7 (Low) · 2026-04-16 · 62%ARC-AGI-2 · Claude 4.7 (Medium) · 2026-04-16 · 68%ARC-AGI-2 · Claude 4.7 (High) · 2026-04-16 · 68%ARC-AGI-2 · Claude 4.7 (Max) · 2026-04-16 · 76%ARC-AGI-2 · GPT-5.5 (Low) · 2026-04-22 · 33%ARC-AGI-2 · GPT-5.5 (Medium) · 2026-04-22 · 70%ARC-AGI-2 · GPT-5.5 (High) · 2026-04-22 · 83%ARC-AGI-2 · GPT-5.5 (XHigh) · 2026-04-22 · 85%ARC-AGI-2 · GPT-5.5 Pro (High) · 2026-04-23 · 85%ARC-AGI-2 · GPT-5.5 Pro (XHigh) · 2026-04-23 · 84%ARC-AGI-2 · Gemini 3.5 Flash (Minimal) · 2026-05-19 · 9%ARC-AGI-2 · Gemini 3.5 Flash (High) · 2026-05-19 · 72%ARC-AGI-2 · Claude Opus 4.8 (Low) · 2026-06-01 · 62%ARC-AGI-2 · Claude Opus 4.8 (Medium) · 2026-06-01 · 72%ARC-AGI-2 · Claude Opus 4.8 (High) · 2026-06-01 · 72%ARC-AGI-2 · Claude Fable 5 (Max) · 2026-06-09 · 89%ARC-AGI-2 · Claude Fable 5 (XHigh) · 2026-06-09 · 88%ARC-AGI-2 · Claude Fable 5 (High) · 2026-06-09 · 88%ARC-AGI-2 · Claude Fable 5 (Medium) · 2026-06-09 · 82%ARC-AGI-2 · Claude Fable 5 (Low) · 2026-06-09 · 77%ARC-AGI-2 · GLM-5.2 · 2026-06-13 · 23%ARC-AGI-2 · GPT-5.6 Sol (Low) · 2026-07-09 · 42%ARC-AGI-2 · GPT-5.6 Sol (Medium) · 2026-07-09 · 67%ARC-AGI-2 · GPT-5.6 Sol (High) · 2026-07-09 · 85%ARC-AGI-2 · GPT-5.6 Sol (XHigh) · 2026-07-09 · 90%ARC-AGI-2 · GPT-5.6 Sol (Max) · 2026-07-09 · 92%ARC-AGI-2 · GPT-5.6 Terra (Low) · 2026-07-09 · 19%ARC-AGI-2 · GPT-5.6 Terra (Medium) · 2026-07-09 · 38%ARC-AGI-2 · GPT-5.6 Terra (High) · 2026-07-09 · 67%ARC-AGI-2 · GPT-5.6 Terra (XHigh) · 2026-07-09 · 74%ARC-AGI-2 · GPT-5.6 Terra (Max) · 2026-07-09 · 84%ARC-AGI-2 · GPT-5.6 Luna (Low) · 2026-07-09 · 5%ARC-AGI-2 · GPT-5.6 Luna (Medium) · 2026-07-09 · 7%ARC-AGI-2 · GPT-5.6 Luna (High) · 2026-07-09 · 29%ARC-AGI-2 · GPT-5.6 Luna (XHigh) · 2026-07-09 · 48%ARC-AGI-2 · GPT-5.6 Luna (Max) · 2026-07-09 · 60%ARC-AGI-2 · Inkling · 2026-07-15 · 37%ARC-AGI-2 · Grok 4.5 (High) · 2026-07-16 · 53%ARC-AGI-2 · Grok 4.5 (Medium) · 2026-07-16 · 53%ARC-AGI-2 · Grok 4.5 (Low) · 2026-07-16 · 33%ARC-AGI-2 · Kimi K3 (Max) · 2026-07-16 · 60%ARC-AGI-2 · Kimi K3 (High) · 2026-07-16 · 55%ARC-AGI-2 · Kimi K3 (Low) · 2026-07-16 · 12%ARC-AGI-2 · Gemini 3.5 Flash-Lite (High) · 2026-07-21 · 10%ARC-AGI-2 · Gemini 3.5 Flash-Lite (Medium) · 2026-07-21 · 5%ARC-AGI-2 · Gemini 3.5 Flash-Lite (Low) · 2026-07-21 · 2%ARC-AGI-2 · Gemini 3.5 Flash-Lite (Minimal) · 2026-07-21 · 1%ARC-AGI-2 · Gemini 3.6 Flash (High) · 2026-07-21 · 60%ARC-AGI-2 · Gemini 3.6 Flash (Medium) · 2026-07-21 · 50%ARC-AGI-2 · Gemini 3.6 Flash (Low) · 2026-07-21 · 30%ARC-AGI-2 · Gemini 3.6 Flash (Minimal) · 2026-07-21 · 3%ARC-AGI-2 · Claude Opus 5 (High) · 2026-07-24 · 88%ARC-AGI-2 · Claude Opus 5 (Max) · 2026-07-24 · 90%ARC-AGI-2 · Inkling Small (XHigh) · 2026-07-30 · 40%ARC-AGI-2 · Inkling Small (High) · 2026-07-30 · 33%ARC-AGI-2 · Inkling Small (Medium) · 2026-07-30 · 14%ARC-AGI-2 · Inkling Small (Low) · 2026-07-30 · 5%ARC-AGI-2 · Inkling Small (Minimal) · 2026-07-30 · 4%ARC-AGI-2 · Inkling Small (None) · 2026-07-30 · 0%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (Max) · 2026-07-30 · 60%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (XHigh) · 2026-07-30 · 43%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (High) · 2026-07-30 · 33%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (Medium) · 2026-07-30 · 7%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (Low) · 2026-07-30 · 4%ARC-AGI-2 · GPT-5.6 Luna 2026-07-30 (None) · 2026-07-30 · 1%ARC-AGI-2 · DeepSeek V4 Flash 0731 (Max) · 2026-07-31 · 61%ARC-AGI-2 · DeepSeek V4 Flash 0731 (High) · 2026-07-31 · 56%ARC-AGI-2 · DeepSeek V4 Flash 0731 (Low) · 2026-07-31 · 46%ARC-AGI-2 · DeepSeek V4 Flash 0731 (None) · 2026-07-31 · 2%ARC-AGI-2 · Grok 4.6 (XHigh) · 2026-08-11 · 67%ARC-AGI-2 · Grok 4.6 (High) · 2026-08-11 · 65%ARC-AGI-2 · Grok 4.6 (Medium) · 2026-08-11 · 61%ARC-AGI-2 · Grok 4.6 (Low) · 2026-08-11 · 28%ARC-AGI-2 · Gemini 3.7 Flash (High) · 2026-08-13 · 85%ARC-AGI-2 · Gemini 3.7 Flash (Medium) · 2026-08-13 · 64%ARC-AGI-2 · Gemini 3.7 Flash (Low) · 2026-08-13 · 53%ARC-AGI-2 · DeepSeek V4 Pro 0813 (Max) · 2026-08-13 · 61%ARC-AGI-2 · DeepSeek V4 Pro 0813 (High) · 2026-08-13 · 60%ARC-AGI-2 · DeepSeek V4 Pro 0813 (Low) · 2026-08-13 · 56%ARC-AGI-2 · DeepSeek V4 Pro 0813 (None) · 2026-08-13 · 1%ARC-AGI-2 · Dots3-Note Preview (Max) · 2026-08-14 · 77%ARC-AGI-2 · Qwen3.8-27B (XHigh) · 2026-08-14 · 42%ARC-AGI-2 · Qwen3.8-27B (Medium) · 2026-08-14 · 13%ARC-AGI-2 · Qwen3.8-27B (Low) · 2026-08-14 · 23%ARC-AGI-2 · Qwen3.8-27B (None) · 2026-08-14 · 2%ARC-AGI-2 · GLM-5.3-Flash (Max) · 2026-08-26 · 66%ARC-AGI-2 · GLM-5.3-Flash (High) · 2026-08-26 · 50%ARC-AGI-2 · GLM-5.3-Flash (Low) · 2026-08-26 · 28%ARC-AGI-2 · Claude Fable 5.1 (Max) · 2026-09-01 · 90%ARC-AGI-2 · Claude Fable 5.1 (XHigh) · 2026-09-01 · 90%ARC-AGI-2 · Claude Fable 5.1 (High) · 2026-09-01 · 89%ARC-AGI-2 · Claude Fable 5.1 (Medium) · 2026-09-01 · 86%ARC-AGI-2 · Claude Fable 5.1 (Low) · 2026-09-01 · 78%ARC-AGI-2 · GPT-6 Astra (Max) · 2026-09-02 · 95%ARC-AGI-2 · GPT-6 Astra (XHigh) · 2026-09-02 · 93%ARC-AGI-2 · GPT-6 Astra (High) · 2026-09-02 · 92%ARC-AGI-2 · GPT-6 Astra (Medium) · 2026-09-02 · 92%ARC-AGI-2 · GPT-6 Astra (Low) · 2026-09-02 · 85%ARC-AGI-2 · GPT-6 Astra (None) · 2026-09-02 · 60%ARC-AGI-2 · Gemini 3.8 Flash (High) · 2026-09-02 · 89%ARC-AGI-2 · Gemini 3.8 Flash (Medium) · 2026-09-02 · 83%ARC-AGI-2 · Gemini 3.8 Flash (Low) · 2026-09-02 · 78%ARC-AGI-2 · DeepSeek V4.1 Flash (Max) · 2026-09-10 · 73%ARC-AGI-2 · DeepSeek V4.1 Flash (High) · 2026-09-10 · 68%ARC-AGI-2 · DeepSeek V4.1 Flash (Low) · 2026-09-10 · 61%ARC-AGI-2 · DeepSeek V4.1 Flash (None) · 2026-09-10 · 1%ARC-AGI-2 · Grok 4.7 (XHigh) · 2026-09-21 · 61%ARC-AGI-2 · Grok 4.7 (High) · 2026-09-21 · 58%ARC-AGI-2 · Grok 4.7 (Medium) · 2026-09-21 · 59%ARC-AGI-2 · Grok 4.7 (Low) · 2026-09-21 · 17%ARC-AGI-2 · GPT-6 Luna (Max) · 2026-09-22 · 59%ARC-AGI-2 · GPT-6 Luna (XHigh) · 2026-09-22 · 42%ARC-AGI-2 · GPT-6 Luna (High) · 2026-09-22 · 31%ARC-AGI-2 · GPT-6 Luna (Medium) · 2026-09-22 · 18%ARC-AGI-2 · GPT-6 Luna (Low) · 2026-09-22 · 5%ARC-AGI-2 · GPT-6 Luna (None) · 2026-09-22 · 0%ARC-AGI-2 · Claude Opus 5.5 (Max) · 2026-09-22 · 92%ARC-AGI-2 · Claude Opus 5.5 (XHigh) · 2026-09-22 · 92%ARC-AGI-2 · Claude Opus 5.5 (High) · 2026-09-22 · 93%ARC-AGI-2 · Claude Opus 5.5 (Medium) · 2026-09-22 · 88%ARC-AGI-2 · Claude Opus 5.5 (Low) · 2026-09-22 · 70%ARC-AGI-2 · GPT-6 Sol (Max) · 2026-09-22 · 90%ARC-AGI-2 · GPT-6 Sol (XHigh) · 2026-09-22 · 78%ARC-AGI-2 · GPT-6 Sol (High) · 2026-09-22 · 69%ARC-AGI-2 · GPT-6 Sol (Medium) · 2026-09-22 · 58%ARC-AGI-2 · GPT-6 Sol (Low) · 2026-09-22 · 32%ARC-AGI-2 · GPT-6 Sol (None) · 2026-09-22 · 2%ARC-AGI-2 · GPT-6.1 Sol (Max) · 2026-09-29 · 94%ARC-AGI-2 · GPT-6.1 Sol (XHigh) · 2026-09-29 · 92%ARC-AGI-2 · GPT-6.1 Sol (High) · 2026-09-29 · 92%ARC-AGI-2 · GPT-6.1 Sol (Medium) · 2026-09-29 · 87%ARC-AGI-2 · GPT-6.1 Sol (Low) · 2026-09-29 · 77%ARC-AGI-2 · Claude Haiku 5.5 (XHigh) · 2026-10-07 · 61%ARC-AGI-2 · Claude Haiku 5.5 (High) · 2026-10-07 · 31%ARC-AGI-2 · Claude Haiku 5.5 (Medium) · 2026-10-07 · 9%ARC-AGI-2 · Claude Haiku 5.5 (Low) · 2026-10-07 · 3%ARC-AGI-2 · Gemini 4 Argon (High) · 2026-10-13 · 91%ARC-AGI-2 · Gemini 4 Argon (Medium) · 2026-10-13 · 87%ARC-AGI-2 · Gemini 4 Argon (Low) · 2026-10-13 · 42%ARC-AGI-2now

Stepped line: best score to date. Points: every reported result. FrontierMath (tiers 1–3) from Epoch AI; ARC-AGI semi-private set from ARC Prize (undocumented endpoint), plotted by model release date. ARC's own human-panel row is drawn as a reference line, as ARC Prize lists it.

Crowd vs. CEOs

Prediction-market odds that AGI arrives before each year (Manifold), set against the windows lab leaders have given. These are separate markets with their own resolution criteria, and none uses a CEO's definition of AGI. Read the line as the crowd's rough timing, not as a calibrated distribution.

202620272028202920302031203220332034203520360%25%50%75%100%P(AGI before year)Will we get AGI before 2027?: 3%Will we get AGI before 2028?: 22%Will we get AGI before 2030?: 50%Will we get AGI before 2032?: 62%Elon Musk · 2024Elon Musk: AI smarter than any one human around end of 2025Dario Amodei · 2025◂ 2025-09Dario Amodei: AI writes 90% of code within 3-6 monthsSam Altman · 2025Sam Altman: Novel-insight systems in 2026Dario Amodei · 2024Dario Amodei: Powerful AI could arrive as early as 2026Sam Altman · 2025Sam Altman: OpenAI target: intern-level AI research assistant by Sept 2026Anthropic · 2025Anthropic: Powerful AI systems (Nobel-level across most disciplines) by late 2026 / early 2027Leopold Aschenbrenner · 2024Leopold Aschenbrenner: AGI by 2027 is strikingly plausibleDario Amodei · 2026Dario Amodei: Powerful AI could be 1-2 years awaySam Altman · 2025Sam Altman: OpenAI target: fully automated AI researcher by 2028Geoffrey Hinton · 2023Geoffrey Hinton: AI may reason better than humans in five yearsJensen Huang · 2024Jensen Huang: AI performs well on every test within five yearsMark Zuckerberg · 2025Mark Zuckerberg: The rest of this decade is decisive for superintelligence's pathDemis Hassabis · 2025Demis Hassabis: AGI within five to ten yearsSam Altman · 2025Sam Altman: By 2035 one person can command the intellectual capacity of all of 2025Daniel Kokotajlo et al. · 2025Daniel Kokotajlo et al.: AI 2027 scenario: superhuman AI impact over the next decade exceeds the Industrial RevolutionDemis Hassabis · 2025Demis Hassabis: AI could end disease within about a decadenow

Observed 2026-10-19 04:51 UTC. Below the line: every dated horizon from the current-claims set, sorted by when it opens and coloured by status.

Superhuman Ledger

The first time an AI system matched or beat the best humans, or expert humans, at something. "Superhuman" is often contested, so every row states its criterion and a confidence grade. Of 31 events, 24 came from industry, 4 from academia and 3 from collaborations.

exceeded best humans matched best exceeded expert median contested hollow = low confidence
19901995200020052010201520202025gamesvisionspeechmedicinebiologylawroboticsearth-sciencemathcoding1994-08-18 · Declared Man-Machine World Checkers Champion · contested · medium confidence1995-03-01 · Backgammon play at or near world-class human level · matched best · medium confidence1997-05-01 · Defeated reigning world chess champion in a standard-time match · exceeded best · high confidence2007-09-14 · Checkers weakly solved: perfect play by both sides is a draw · exceeded best · high confidence2011-02-01 · Defeated the two best Jeopardy! champions in a televised match · exceeded best · high confidence2015-02-06 · First reported ImageNet classification top-5 error below cited human error · contested · medium confidence2016-01-27 · Beat a professional Go player 5-0 in full tournament conditions · exceeded expert median · medium confidence2016-03-01 · Defeated Lee Sedol 4-1 in Seoul · exceeded best · high confidence2016-10-17 · Reported human parity on conversational speech recognition · contested · medium confidence2017-01-06 · Defeated professional poker players in heads-up no-limit hold'em · exceeded expert median · medium confidence2017-01-30 · Beat four top heads-up no-limit hold'em professionals over 120000 hands · exceeded best · high confidence2017-10-18 · Self-play-only system beat the version that defeated Lee Sedol 100-0 · exceeded best · high confidence2017-11-14 · Pneumonia detection on chest X-ray exceeds average radiologist F1 · contested · medium confidence2017-12-05 · Single algorithm superhuman at chess shogi and Go from rules alone · exceeded best · medium confidence2019-04-13 · Defeated Dota 2 world champions Team OG · exceeded best · medium confidence2019-07-11 · Beat elite professionals in six-player no-limit hold'em · exceeded best · high confidence2019-10-30 · Grandmaster level in StarCraft II in all three races · contested · medium confidence2020-03-30 · First deep RL agent above the standard human benchmark on all 57 Atari games · exceeded expert median · medium confidence2020-11-30 · Median GDT 92.4 at CASP14 near experimental accuracy · matched best · medium confidence2022-02-09 · Outraced top Gran Turismo e-sports drivers · exceeded best · medium confidence2022-11-22 · Human-level play in Diplomacy with natural language negotiation · exceeded expert median · medium confidence2023-03-15 · Simulated bar exam score in roughly the top 10% of test takers · contested · medium confidence2023-08-30 · Autonomous drone matched/beat human FPV champions in head-to-head races · matched best · high confidence2023-11-14 · Beat the leading operational deterministic forecast on most targets · exceeded best · high confidence2024-07-25 · Silver-medal-level score at the IMO (28/42) · exceeded expert median · high confidence2025-07-19 · Gold-medal-level score at the IMO (35/42) in natural language within time limit · exceeded expert median · medium confidence2025-07-21 · Gold-medal-standard score at the IMO (35/42) in natural language within time limit · exceeded expert median · high confidence2025-08-01 · Gold-level score at the IOI online track (6th of 330) · exceeded expert median · low confidence2025-09-04 · Gold-level result at the ICPC World Finals (10/12 problems) · exceeded expert median · high confidence2025-09-04 · Solved all 12 problems at ICPC World Finals · exceeded best · medium confidence2026-09-02 · First AI to outscore the top human on an IOI problem set (535.4/600 vs 498.27) · exceeded best · medium confidence
DateAchievementCriterionLevel
2026-09-02 First AI to outscore the top human on an IOI problem set (535.4/600 vs 498.27)
Nemotron-3-Ultra-CC competition-specific · NVIDIA
Score on IOI 2026 problems under same time and submission constraints exceeded best
medium confidence
2025-09-04 Solved all 12 problems at ICPC World Finals
GPT-5 + experimental reasoning model · OpenAI
12 of 12 problems vs best human team 11 exceeded best
medium confidence
2025-09-04 Gold-level result at the ICPC World Finals (10/12 problems)
Gemini 2.5 Deep Think · Google DeepMind
Solved 10 of 12 problems; would rank 2nd by time (677 min); solved Problem C no team solved exceeded expert median
high confidence
2025-08 Gold-level score at the IOI online track (6th of 330)
Ensemble of general-purpose reasoning models · OpenAI
Ranked 6th overall under 5-hour limit and 50-submission cap exceeded expert median
low confidence
2025-07-21 Gold-medal-standard score at the IMO (35/42) in natural language within time limit
Gemini Deep Think · Google DeepMind
Solved 5 of 6 problems for 35 of 42 points; graded by IMO coordinators exceeded expert median
high confidence
2025-07-19 Gold-medal-level score at the IMO (35/42) in natural language within time limit
Experimental reasoning LLM (unreleased) · OpenAI
Solved 5 of 6 problems for 35 of 42 points; graded by three former IMO medalists (unanimous consensus), per the researcher post exceeded expert median
medium confidence
2024-07-25 Silver-medal-level score at the IMO (28/42)
AlphaProof + AlphaGeometry 2 · Google DeepMind
Solved 4 of 6 problems for 28 of 42 points exceeded expert median
high confidence
2023-11-14 Beat the leading operational deterministic forecast on most targets
GraphCast · Google DeepMind
Outperformed HRES on over 90% of 1380 verification targets exceeded best
high confidence
2023-08-30 Autonomous drone matched/beat human FPV champions in head-to-head races
Swift · University of Zurich / Intel
Won several races vs each champion and recorded fastest race time matched best
high confidence
2023-03-15 Simulated bar exam score in roughly the top 10% of test takers
GPT-4 · OpenAI
Percentile on simulated bar exam contested
medium confidence
2022-11-22 Human-level play in Diplomacy with natural language negotiation
CICERO · Meta AI
Over 40 games: more than 2x average human score and top 10% of multi-game players exceeded expert median
medium confidence
2022-02-09 Outraced top Gran Turismo e-sports drivers
Gran Turismo Sophy · Sony AI
Head-to-head competition vs top drivers exceeded best
medium confidence
2020-11-30 Median GDT 92.4 at CASP14 near experimental accuracy
AlphaFold 2 · Google DeepMind
Median GDT across targets vs informal threshold of ~90 matched best
medium confidence
2020-03-30 First deep RL agent above the standard human benchmark on all 57 Atari games
Agent57 · DeepMind
Score above human baseline on every game of the 57-game suite exceeded expert median
medium confidence
2019-10-30 Grandmaster level in StarCraft II in all three races
AlphaStar · Google DeepMind
Ranked above 99.8% of active players contested
medium confidence
2019-07-11 Beat elite professionals in six-player no-limit hold'em
Pluribus · Meta AI / Carnegie Mellon University
10000 hands 5 humans+1 AI over 12 days; 5000 hands each for 3 pros vs 5 AIs exceeded best
high confidence
2019-04-13 Defeated Dota 2 world champions Team OG
OpenAI Five · OpenAI
Won match vs reigning champions exceeded best
medium confidence
2017-12-05 Single algorithm superhuman at chess shogi and Go from rules alone
AlphaZero · Google DeepMind
Within 24 hours reached superhuman level; defeated world-champion program in each game exceeded best
medium confidence
2017-11-14 Pneumonia detection on chest X-ray exceeds average radiologist F1
CheXNet · Stanford University
F1 on pneumonia detection vs average of 4 radiologists contested
medium confidence
2017-10-18 Self-play-only system beat the version that defeated Lee Sedol 100-0
AlphaGo Zero · Google DeepMind
100-0 vs AlphaGo Lee after 3 days; 'quickly surpassed human level' exceeded best
high confidence
2017-01-30 Beat four top heads-up no-limit hold'em professionals over 120000 hands
Libratus · Carnegie Mellon University
Led pros by $1766250 in chips over 20 days; developers state statistically significant exceeded best
high confidence
2017-01-06 Defeated professional poker players in heads-up no-limit hold'em
DeepStack · University of Alberta / Charles University / CTU Prague
Win rate with statistical significance over 44000 hands exceeded expert median
medium confidence
2016-10-17 Reported human parity on conversational speech recognition
Microsoft conversational speech system · Microsoft Research
WER 5.8% vs 5.9% Switchboard; 11.0% vs 11.3% CallHome contested
medium confidence
2016-03 Defeated Lee Sedol 4-1 in Seoul
AlphaGo · Google DeepMind
Match result 4-1 exceeded best
high confidence
2016-01-27 Beat a professional Go player 5-0 in full tournament conditions
AlphaGo · Google DeepMind
5-0 match exceeded expert median
medium confidence
2015-02-06 First reported ImageNet classification top-5 error below cited human error
PReLU-Net · Microsoft Research
Top-5 error 4.94% vs 5.1% on ImageNet 2012 classification contested
medium confidence
2011-02 Defeated the two best Jeopardy! champions in a televised match
Watson · IBM
Final score $77147 vs $24000 and $21600 exceeded best
high confidence
2007-09-14 Checkers weakly solved: perfect play by both sides is a draw
Chinook · University of Alberta
Game-theoretic solution (cannot be beaten) exceeded best
high confidence
1997-05 Defeated reigning world chess champion in a standard-time match
Deep Blue · IBM
Won 6-game rematch 3.5-2.5 vs reigning champion exceeded best
high confidence
1995-03 Backgammon play at or near world-class human level
TD-Gammon 2.1 · IBM Research (G. Tesauro)
Net points in exhibition sessions vs top players; expert assessment matched best
medium confidence
1994-08-18 Declared Man-Machine World Checkers Champion
Chinook · University of Alberta
Title awarded by organizers after Tinsley resigned the match (six games all drawn) citing health; no match win over Tinsley contested
medium confidence

All claims

Quotes are verbatim from the linked source. Every quote is re-checked weekly against the live page (last check 2026-10-16); a quote that can no longer be found there is flagged, not hidden. 3 further claims are held off the page until the quote is confirmed at its source.

Current: 2020 onward

SaidClaimHorizonStatus
2026-01 Dario Amodei CEO, Anthropic
Powerful AI could be 1-2 years away
“could be as little as 1-2 years away”
source (primary) · quote check: verbatim
2027-2028 open
2025-10-28 Sam Altman CEO, OpenAI
OpenAI target: intern-level AI research assistant by Sept 2026
“an intern-level research assistant by September 2026”
source (press) · quote check: verbatim
2026-09 deadline passed: outcome unverified
2025-10-28 Sam Altman CEO, OpenAI
OpenAI target: fully automated AI researcher by 2028
“a fully automated 'legitimate AI researcher' by 2028”
source (press) · quote check: verbatim
2028 open
2025-07-30 Mark Zuckerberg CEO, Meta
Superintelligence development is 'now in sight'
“Developing superintelligence is now in sight.”
source (official) · quote check: verbatim
— undated
2025-07-30 Mark Zuckerberg CEO, Meta
The rest of this decade is decisive for superintelligence's path
“The rest of this decade seems likely to be the decisive period for determining the path this technology will take”
source (official) · quote check: verbatim
2030 open
2025-04-20 Demis Hassabis CEO, Google DeepMind
AGI within five to ten years
“In the next five to ten years, I think.”
source (press) · quote check: verbatim
2030-2035 open
2025-04-20 Demis Hassabis CEO, Google DeepMind
AI could end disease within about a decade
“Maybe within the next decade or so, I don't see why not.”
source (press) · quote check: verbatim · Answer to the interviewer's question "The end of disease?"
2035 open
2025-04-03 Daniel Kokotajlo et al. AI Futures Project
AI 2027 scenario: superhuman AI impact over the next decade exceeds the Industrial Revolution
“We predict that the impact of superhuman AI over the next decade will be enormous”
source (primary) · quote check: verbatim
2035 open
2025-03-10 Dario Amodei CEO, Anthropic
AI writes 90% of code within 3-6 months
“we are not far from the world—I think we'll be there in three to six months—where AI is writing 90 percent of the code.”
source (official) · quote check: verbatim
2025-09 deadline passed: outcome unverified
2025-03-06 Anthropic AI lab, submission to US OSTP
Powerful AI systems (Nobel-level across most disciplines) by late 2026 / early 2027
“we expect powerful AI systems will emerge in late 2026 or early 2027.”
source (official) · quote check: verbatim
2027 open
2025 Sam Altman CEO, OpenAI
OpenAI knows how to build AGI; aiming beyond it to superintelligence
“We are now confident we know how to build AGI as we have traditionally understood it.”
source (primary) · quote check: verbatim
— undated
2025 Sam Altman CEO, OpenAI
By 2035 one person can command the intellectual capacity of all of 2025
“Anyone in 2035 should be able to marshall the intellectual capacity equivalent to everyone in 2025”
source (primary) · quote check: verbatim
2035 open
2025 Sam Altman CEO, OpenAI
Novel-insight systems in 2026
“2026 will likely see the arrival of systems that can figure out novel insights.”
source (primary) · quote check: verbatim
2026 window open
2024-10 Dario Amodei CEO, Anthropic
Powerful AI could arrive as early as 2026
“I think it could come as early as 2026, though there are also ways it could take much longer.”
source (primary) · quote check: verbatim
2026 window open
2024-10 Dario Amodei CEO, Anthropic
Powerful AI compresses 50-100 years of biology progress into 5-10 years
“compress the progress that human biologists would have achieved over the next 50-100 years into 5-10 years”
source (primary) · quote check: verbatim
— undated
2024-09-23 Sam Altman CEO, OpenAI
Superintelligence in 'a few thousand days'
“It is possible that we will have superintelligence in a few thousand days”
source (primary) · quote check: verbatim
— undated
2024-06 Leopold Aschenbrenner Ex-OpenAI; Situational Awareness
AGI by 2027 is strikingly plausible
“AGI by 2027 is strikingly plausible.”
source (primary) · quote check: verbatim
2027 open
2024-04-09 Elon Musk CEO, Tesla/xAI
AI smarter than any one human around end of 2025
“My guess is that we'll have AI that is smarter than any one human probably around the end of next year,”
source (press) · quote check: verbatim
2025 deadline passed: outcome unverified
2024-03-07 Yann LeCun Chief AI Scientist, Meta (at the time)
Skeptic: generative/LLM approach will not reach human-level AI
“If you're really interested in human level AI, abandon the idea of generative AI.”
source (press) · quote check: verbatim · Date is when the transcript was posted, which may differ from the recording date.
— undated
2024-03 Jensen Huang CEO, Nvidia
AI performs well on every test within five years
“and I'm guessing in five years time, we'll do well on every single one,”
source (press) · quote check: verbatim
2029 open
2023-10-08 Geoffrey Hinton Univ. of Toronto; Nobel laureate
AI may reason better than humans in five years
“I think in five years' time it may well be able to reason better than us.”
source (press) · quote check: verbatim
2028 open

Graveyard: historical forecasts

SaidClaimHorizonStatus
2024-01 Katja Grace et al. AI Impacts
2023 survey: 10% by 2027, 50% by 2047 for machines outperforming humans on every task
“If science continues undisrupted, the chance of unaided machines outperforming humans in every possible task was estimated at 10% by 2027, and 50% by 2047.”
source (paper) · quote check: verbatim
2047 open
2022-08 Zach Stein-Perlman, Katja Grace et al. AI Impacts
2022 survey: 50% chance of HLMI by 2059
“The aggregate forecast time to a 50% chance of HLMI was 37 years, i.e. 2059”
source (paper) · quote check: verbatim
2059 open
2017-05 Katja Grace et al. AI Impacts / Oxford / Yale
Expert survey: 50% chance AI outperforms humans in all tasks in 45 years; all jobs automated in 120 years
“Researchers believe there is a 50% chance of AI outperforming humans in all tasks in 45 years and of automating all human jobs in 120 years”
source (paper) · quote check: verbatim
2061 open
2014 Vincent C. Muller and Nick Bostrom Future of Humanity Institute, Oxford
Expert survey (2012/13): 50% chance of HLMI around 2040-2050; 90% by 2075
“The median estimate of respondents was for a one in two chance that high-level machine intelligence will be developed around 2040-2050, rising to a nine in ten chance by 2075.”
source (paper) · quote check: verbatim
2040-2050 open
2002 Ray Kurzweil (Long Bets wager vs Mitchell Kapor) Inventor and author
A machine will pass the Turing test by 2029
“Ray Kurzweil maintains that a computer (i.e., a machine intelligence) will pass the Turing test by 2029.”
source (primary) · quote check: verbatim
2029 open
1998 Hans Moravec Carnegie Mellon Univ.
Hardware matching the human brain available in cheap machines by the 2020s
“At the present rate, computers suitable for humanlike robots will appear in the 2020s.”
source (paper) · quote check: verbatim
2020s window open
1993-03 Vernor Vinge San Diego State Univ.
Technological means for superhuman intelligence within 30 years
“Within thirty years, we will have the technological means to create superhuman intelligence.”
source (primary) · quote check: verbatim
2023 deadline passed: outcome unverified
1993-03 Vernor Vinge San Diego State Univ.
The singularity will occur between 2005 and 2030
“I'll be surprised if this event occurs before 2005 or after 2030.”
source (primary) · quote check: verbatim
2005-2030 window open
1973 Sir James Lighthill Univ. of Cambridge; report to UK Science Research Council
Retrospective: AI had not delivered the promised impact (the Lighthill report preceding the first AI winter)
“In no part of the field have the discoveries made so far produced the major impact that was then promised”
source (secondary) · quote check: verbatim · A retrospective assessment, not a forecast. Quoted via Wikipedia; the original publication was not consulted.
— undated
1970 Marvin Minsky (as reported by Life magazine) MIT
General-intelligence machine within 3-8 years
“In from three to eight years we will have a machine with the general intelligence of an average human being.”
source (secondary) · quote check: verbatim · Attribution caution: the wording is from a 1970 Life magazine article by a journalist, not Minsky's own writing.
1978 deadline passed: outcome unverified
1967 Marvin Minsky MIT
AI will be substantially solved within a generation
“Within a generation... the problem of creating 'artificial intelligence' will substantially be solved.”
source (secondary) · quote check: verbatim · Quoted via Wikipedia; the original publication was not consulted. 'A generation' gives no date, so no horizon is recorded.
— undated
1965 Herbert A. Simon Carnegie Tech
Machines will do any work a man can do within twenty years
“machines will be capable, within twenty years, of doing any work a man can do.”
source (secondary) · quote check: verbatim · Quoted via Wikipedia; the original publication was not consulted.
1985 deadline passed: outcome unverified
1965 I. J. Good Virginia Tech
The first ultraintelligent machine would trigger an intelligence explosion and be humanity's last necessary invention
“the first ultraintelligent machine is the last invention that man need ever make”
source (secondary) · quote check: verbatim · Quoted via Wikipedia; the original publication was not consulted.
— undated
1958 Herbert A. Simon and Allen Newell Carnegie Tech
A computer will be world chess champion within ten years
“within ten years a digital computer will be the world's chess champion”
source (secondary) · quote check: verbatim · Quoted via Wikipedia; the original publication was not consulted.
1968 achieved
1997, +28 yr vs. horizon
1958 Herbert A. Simon and Allen Newell Carnegie Tech
A computer will discover and prove an important new theorem within ten years
“within ten years a digital computer will discover and prove an important new mathematical theorem.”
source (secondary) · quote check: verbatim · Quoted via Wikipedia; the original publication was not consulted.
1968 deadline passed: outcome unverified
1955-08-31 John McCarthy et al. Dartmouth summer project proposal
A carefully chosen group working together for a summer could make significant advances on machine intelligence problems
“We think that a significant advance can be made in one or more of these problems”
source (primary) · quote check: verbatim
— undated
1950 Alan Turing Univ. of Manchester; Mind
By ~2000 a computer will hold an average interrogator to <=70% correct identification after 5 minutes
“I believe that in about fifty years' time it will be possible, to programme computers, with a storage capacity of about 109, to make them play the imitation game so well that an average interrogator will not have more than 70 per cent chance of making the right identification after five minutes of questioning.”
source (paper) · quote check: verbatim
2000 deadline passed: outcome unverified
1950 Alan Turing Univ. of Manchester; Mind
By ~2000 educated opinion will accept the phrase 'machines thinking'
“Nevertheless I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted.”
source (paper) · quote check: verbatim
2000 deadline passed: outcome unverified