Changes, page 10 of 37

Every substantive change to the register, newest first. Speaker events record what the people covered said or changed. Site events record what this site published, corrected, flagged, or withdrew. The two never share a label.

Substantive changes (7383)

Page 10 of 37: changes 1801 to 2000 of 7383, newest first. Dates are when the change was recorded.

2026-10-29
Assessment publishedBy this siteEd Zitron
Outside of ChatGPT there is really no proof that generative AI products are remotely essential, and he argues there is very little about ChatGPT that Microsoft could not provide, with rate limits, through Copilot.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-is-a-money-trap-04963c-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-fff2d9
a conditional_forecast claim cannot be assessed on the historical_observation axis
2026-10-29
Assessment publishedBy this siteEd Zitron
The AI bubble has every problem that burst the Dot-Com Bubble, in worse and bigger form, with global contagion risk.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-f9d8c2-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-f3e4c9
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-f2344c
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
NVIDIA appears to be channel stuffing by pushing customers to take early delivery of GPUs.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-ef930f-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-e636e6
The openai review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-e24951
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-e23af3
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-c8bc6c
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-c854ec
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-c54db0
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-c4e60e
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-c3051e
e1 measures something else (metric Three direct customers' combined share of first-half FY2027 revenue is not Direct-customer concentration in first-half and Q2 FY2027 revenue, together with Q2-end accounts-receivable concentration); it can only be context unless it is derived with stated inputs; e2 measures something else (metric Direct-customer concentration in accounts receivable at Q2 FY2027 end is not Direct-customer concentration in first-half and Q2 FY2027 revenue, together with Q2-end accounts-receivable concentration); it can only be context unless it is derived with stated inputs
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-be0b15
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-bcb3a4
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-b7d4fe
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-aa12f8
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-a82587
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteEd Zitron
Generative AI has made everything more annoying, expensive and worse, including US Treasury prices.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-907fa0-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-8fc147
e3 measures something else (period 2026-09-01..2026-09-30 is not 2026-09-01..2026-10-20); it can only be context unless it is derived with stated inputs
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-7cb19e
The final-text check did not complete (unavailable).
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-6ba64c
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
The elevated valuations of AI-related stocks in Taiwan, Japan and Korea come from a belief they will grow in perpetuity.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-684909-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The consequences of the AI bubble are global and at a scale unseen in history, larger than the Dot-Com Bubble.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-62dc6e-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-5f1a93
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteEd Zitron
The vast majority of AI demand comes from selling services that do not make any economic sense.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-5a1683-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-598363
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-55866b
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteEd Zitron
If people lose faith in the AI trade, everything will start to collapse.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-4bd256-a2.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-47ca97
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
NVIDIA is 11.31% of the NASDAQ 100, and nine AI-trade companies make up about 46.5% of the NASDAQ 100 and 39.3% of the NASDAQ Composite.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-3a3757-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The AI bubble has driven the highest inflow of foreign investment into US equities in history, larger than the Dot-Com era, pre-GFC, or 2021.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-39d19e-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The US government cannot bail out future misses of analyst expectations by NVIDIA, Microsoft, Google, Amazon, CoreWeave or Meta, cannot cover every single bankrupt data center, and cannot bail out investor sentiment once investors realise the AI boom was another tech bust.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-357bf7-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
If hyperscaler and neocloud demand wavers, NVIDIA, Broadcom, Micron, AMD and related suppliers will see revenue growth stumble and their stocks suffer badly.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-2b2867-a2.
2026-10-29
Assessment publishedBy this siteEd Zitron
Sustaining AI-trade stock valuations requires every company involved to spend or make hundreds of billions of dollars in perpetuity, increasing every quarter.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-1dc891-a2.
2026-10-29
Assessment publishedBy this siteEd Zitron
Venture capital is not particularly good at creating value and has put the majority of its investments into a single industry (AI), with the majority of those going to LLM-related or LLM-powered companies that have yet to prove their economic worth.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-1c4a00-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-1bb759
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-18d11f
The anthropic review does not accept e1 as contradicts.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-15c793
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-14cfa8
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-0e0485
The openai review doubts the attribution.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-08d781
The two reviews differ: anthropic causal_assessment/not_applicable/mixed, openai causal_assessment/not_applicable/unresolved.
2026-10-29
Assessment publishedBy this siteEd Zitron
The AI bubble is too large and unwieldy for things to 'work out,' and Zitron believes 'we have a problem.'
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-had-better-work-out-021ac2-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-had-better-work-out-01b9f3
The final-text check did not complete (unavailable).
2026-10-29
Assessment publishedBy this siteEd Zitron
Sam Altman's claims about 'intelligence too cheap to meter' were lies.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-e4076f-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
AI costs more today than it did three years ago and is not getting cheaper.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-e16d98-a2.
2026-10-29
Assessment publishedBy this siteEd Zitron
There is no evidence that AI labs running their own infrastructure makes anything profitable.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-d0cb00-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
AI needs to work every time without fail and be absolutely flawless and autonomous.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-c53426-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-bd2eee
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
The AI cost problem surfaced in the last few months because Anthropic and OpenAI quietly moved all their enterprise customers to token-based billing in Q1 2026, and because CEOs had encouraged or incentivized workers to use AI as much as possible, in some cases making AI use a job-threatening KPI.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-b3fd43-a2.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-b117d4
The anthropic review does not accept e1 as supports.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-a0de59
The anthropic review reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
The fact that AI companies, backed by the full manufactured consent of the markets and most of the economy, still had to subsidize their products shows exactly how flimsy their value is.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-93d983-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Anyone who relies on LLMs for anything is a mark.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-88b310-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Microsoft intentionally hid the actual cost of GitHub Copilot prompts and let users overspend in order to boost growth.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-865cce-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Every LLM prompt, project and interaction can go wrong in ways that are hard to predict or plan for, short of constant vigilance against catastrophically stupid behaviour, because LLMs have no thoughts, consciousness or ability to learn outside pre-training and post-training.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-683e97-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Anthropic, OpenAI and every other AI company deliberately obscured the costs of AI use because they knew that as soon as users had to pay for a model's mistakes they would scream in protest.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-5de67a-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
OpenAI, Anthropic and other AI companies actively conspired to mislead the world about AI's true costs, and this worked well until they decided to try charging what AI actually costs.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-57410c-a2.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-571465
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteEd Zitron
AI may have had a negative economic impact, raising operating expenses and making some software engineers worse at writing software.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-555c71-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
AI companies did not raise hundreds of billions of dollars by talking about any product currently on the market, because current products are not very good.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-4ca1dd-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The threat of AI job loss allowed Sam Altman and Dario Amodei to raise hundreds of billions of dollars while lying about it, and both have since walked back their job-loss claims.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-46a9e5-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-464112
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteEd Zitron
Every AI startup is inherently unprofitable because they are literally sending every penny of their venture capital directly to Anthropic and OpenAI to power their unprofitable services.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-457c99-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The AI bubble could only be inflated by misleading the public and investors about the costs and efficacy of AI models.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-21ba72-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The overwhelming majority of AI users are completely divorced from, and actively trained to ignore, the true cost of AI tokens, so they naturally use these services in an actively uneconomical way.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-1f38b0-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-ai-doesnt-have-roi-1ba0df
The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: Reuters' September 29, 2025 article reports OpenAI's $8.5 billion cash-burn target for 2025, citing reporting Reuters could not independently verify [e2]. | e1 explanation: "...and cautions against equating continuing errors with no improvement.".
2026-10-29
Assessment publishedBy this siteEd Zitron
Anthropic lets subscribers burn $8 to $13.50 in tokens per subscription dollar, and there is no proof beyond Amodei's theoretical scenarios that Anthropic is profitable on inference.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-ai-doesnt-have-roi-1ae155-a2.
2026-10-29
Assessment publishedBy this siteEd Zitron
Executives are excited about generative AI because they are cut off from production and do not create anything themselves.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-absentee-capitalism-4677ea-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
The WGA's strike demands, including guarantees on how AI is used in union productions, are reasonable.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-absentee-capitalism-24254f-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-fbaeb8
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-fba26b
The two reviews differ: anthropic historical_observation/not_applicable/unresolved, openai not_applicable/not_applicable/not_adjudicable.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-f7b7ea
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-ec9552
hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteEd Zitron
No generative AI company is profitable.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-b4174f-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
OpenAI burned more than $4 billion more cash than had been publicly known.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-a77826-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
AI revenues will never pay back the cost of these GPUs, meaning the NVIDIA GPUs being bought for AI.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-9d1e73-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
CoreWeave, an AI compute neocloud backed by NVIDIA, is a time bomb.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-93fc07-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
NVIDIA's future relies on endless debt, millions of GPUs are sitting waiting to be installed, and it no longer makes sense to buy more GPUs.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-93c46d-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-922795
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
There is no AI SaaS story, and everything in the AI industry rides, and dies, on selling GPUs.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-728c14-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Unprofitable, unsustainable AI threatens the future of venture capital, private equity, and the markets.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-6145c8-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
Data center developers have no exit route, and US economic growth is carried entirely by big tech.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-5fdc35-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
GPT-5's router mode costs OpenAI more money to run.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-5b6fc1-a1.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-4b043f
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteEd Zitron
Held record wyea-2025-a-retrospective-44fd6b
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteEd Zitron
As of September 12, 2025, OpenAI couldn't afford to pay Oracle and Oracle didn't have the capacity to service their $300 billion, five-year deal.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-38aee6-a1.
2026-10-29
Assessment publishedBy this siteEd Zitron
DeepSeek scared the American AI industry because of cost, not because of national security.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-29fe39-a2.
2026-10-29
Assessment publishedBy this siteEd Zitron
OpenAI's revenues were around $4.5 billion by the end of September 2025, a vast difference from the $4.3 billion for the first half of the year published by other outlets.
First assessment published after both reviews and the final-text check passed.
Assessment wyea-2025-a-retrospective-088e29-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
There are few regulations on frontier AI systems, and this may not change soon.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-fc9fb5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Safely building AGI should be treated as a matter of global security, not national security.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-e20c36-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record obsolete-pub-cp-150648474-853efd
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
The timing of Brundage's exit may be connected to OpenAI's attempt to restructure as a for-profit PBC.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-843bd5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Brundage's departure is another sign that the original OpenAI no longer exists.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-729b9f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record obsolete-pub-cp-150648474-538048
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
We should be very wary of letting AI technologists with simplistic world models dictate international relations.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-52174e-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record obsolete-pub-cp-150648474-5141f6
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record obsolete-pub-cp-150648474-393196
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Believing in a clash-of-civilizations race for AGI could motivate profoundly destabilizing actions such as a nuclear first strike.
First assessment published after both reviews and the final-text check passed.
Assessment obsolete-pub-cp-150648474-28bdd6-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record obsolete-pub-cp-150648474-169081
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-e9e901
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
NHTSA data show even ordinary circumstances like avoiding stopped vehicles remain problematic for automated driving systems.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-de30d0-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The new NHTSA data should be a bit of a wakeup call to those who have gone long on Tesla thinking a Full Self Driving solution is in sight; the data aren't devastating, but they aren't so promising either, and they raise a lot of questions.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-ca8bc2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Society should not give up on building driverless cars.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-c509c9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Driverless car experiments should be strictly regulated, and data transparency requirements should be vastly stricter.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-c0396b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Much more data, such as miles driven, is needed if we are to assess in a truly scientific way how safe driver-assist and self-driving systems really are.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-bc7108-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Level 5 self-driving that handles everything has not been achieved.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-b923eb-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Tesla might not survive as a company if its massive bet on self-driving fails.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-ae62e8-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-9ca231
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Without a solution to the outlier problem, full self-driving will not be achieved.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-9ca231-2-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-9b13cd
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-7ea600
hedged language is not scored, even on a plain reading
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-5fcaec
The final-text check failed: rationale_supported_by_evidence; overclaims: AFB's June 2026 report, based on fieldwork ending October 15, 2025, records accessibility and affordability barriers in a survey covering different technologies, not exclusively driverless rides [e3]. | The accessibility survey combines technologies and self-reported experiences; its findings cannot be attributed solely to driverless services [e3]..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-4475d0
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Although he does not know whether it is good or bad, Tesla's roughly 0.75 ADAS incidents per day is enough to be concerning because it comes from a very limited trial of computer-assisted driving relative to the ambition of full self-driving.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-30e65b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the whole US fleet used current self-driving software, incidents would rise with more vehicles and harder conditions, and likely jump radically without humans in the loop.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-25b514-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Vigorous mechanisms should be enforced to ensure drivers using Level 2 driver-assist don't tune out.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-1eaf00-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Companies working on full (Level 5) self-driving are having plenty of problems, per NHTSA ADS data.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-tesla-ever-solve-full-self-driving-1d2f2b-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-tesla-ever-solve-full-self-driving-127663
The final-text check failed: proposition_faithful.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-will-generative-ai-do-more-harm-than-e272e6
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI's negative effects on the employment market have so far been more modest than some expected, while its IP theft has harmed artists and writers.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-generative-ai-do-more-harm-than-d577cd-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Right now (October 2026), the case against generative AI, the ledger of its harms, is quite serious.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-generative-ai-do-more-harm-than-ab933a-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI may pull down the entire economy.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-generative-ai-do-more-harm-than-9f03fd-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
At this point it is not hard to make the argument that generative AI has on the whole been a net detriment to society, although some of his friends have gotten a lot out of it, mainly for coding.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-generative-ai-do-more-harm-than-9edd35-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Turbulent times lie ahead regarding generative AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-will-generative-ai-do-more-harm-than-63ffab-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because everybody competing to build AI is building essentially the same technical solution with essentially the same data, there is no moat for any of them.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-f8412c-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
If there are no clear winners among AI competitors, nobody will be able to charge monopoly prices, and instead there will be price wars and commodity pricing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-e0d84f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Retail investors and retirement funds are likely to be taken advantage of in the AI financing boom.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-ad50c2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
As with SpaceX, Anthropic and OpenAI will be pushed into indices and index trackers and retail investors will suffer losses when reality catches up.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-a036d2-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic and similar companies are ending all-you-can-eat pricing, which will make many customers think twice about using them.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-6ac16d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The economics of AI don't make sense, and more and more people are starting to notice, though it is hard to time when all this might fall apart.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-65c5f2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-things-will-eventually-fall-apart-5cca50
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-things-will-eventually-fall-apart-55bbcb
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
2026 will be remembered as the year retail investors and index funds were left holding the bag on AI investments, though the timing of the collapse is unknown.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-26532e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If there is no moat, no AI company will capture 90% of the AI market.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-things-will-eventually-fall-apart-1acc26-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-openai-may-well-be-completely-ee9022
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's only hope of survival is raising more cash, and it will try to.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-e2f1d9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-openai-may-well-be-completely-bee14e
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI may fail to find a path to profitability because Meta gives away similar technology for free, and this may be fatal to the company.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-9906d6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs have become a commodity available for free.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-810626-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Investors should press OpenAI on its moat, unique technology, killer app, reliability, demo versus reality, and path to profitability.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-6febf8-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus has long suspected that OpenAI was losing a lot of money, though until The Information's analysis he had not seen an analysis showing it.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-60062e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If The Information's analysis is roughly right, OpenAI's losses explain why it reports only revenue and not costs or profits.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-47371c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Customers can easily switch LLM providers, and growing cheaper alternatives from Anthropic and Google forced OpenAI to cut prices in July 2024.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-3d064d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic and several other LLM companies may be fatally undercut by Meta's free models.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-openai-may-well-be-completely-3add89-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-openai-may-well-be-completely-2acc7b
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-f9a4f1
The anthropic review does not accept e1 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
Drones and even nuclear weapons may eventually be placed under LLM command.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-f15bba-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI requires liability, auditing, malpractice standards, and international treaties.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-eed42e-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-e72ad8
hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteGary Marcus
xAI cannot get its house in order even in basic ways.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-e02ba0-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
xAI is in over its head.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-de6bac-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
As of May 2025, Marcus believed actual malice on the part of AI was unlikely any time soon.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-d8f800-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus does not see how building alignment through training data alone can ever work unless systems are cognitively rich enough to represent and reason about moral principles.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-c8c2ff-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-bdbadc
The final-text check failed: citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
Most AI misuse scenarios by bad actors could eventually be stopped, perhaps causing much damage but not extinguishing humanity.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-b94884-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-b0e9b5
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
There is a real risk, especially under the current anti-regulatory regime, that some exceptionally powerful person (Elon or someone else), unconstrained in conventional ways and recklessly disregarding humanity, could accidentally launch and spread an AI that, deliberately or otherwise, causes 'significant harm to the world'.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-aad051-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus puts the probability of AI-caused doom at about 3%, up sharply from a month earlier, while still betting humans prevail for the foreseeable future.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-a36bd6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Industry-standard AI safety measures are unlikely to be sufficient.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-a05140-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus's p(doom) is lower than that of most AI industry people.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-961123-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-94bd02
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus agrees with Samuel Marks's thread that xAI's conduct in releasing Grok 4 without following the industry's safety conventions was way out of line.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-8cb69b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
From xAI's account of the incident, Marcus infers that xAI's main methodology for handling alignment is trial-and-error.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-881381-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is genuinely scary that Elon Musk went from being one of the first industry leaders to warn about AI risks to being the most reckless of the AI leaders.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-505f15-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs should not be put into military systems.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-44a2e9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
We should be wary of forcing unreliable AI throughout decision-making chains.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-41f513-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus says it became clear to him over the last few days that Elon Musk is a person who matches his list of things one might need to really screw up the planet.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-3ace06-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-345f9c
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Grok is so bound to Musk that it sometimes searches for Musk's opinion before answering, and in one case even asked his opinion on pizza.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-2e8c8b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Serious AI regulation is needed because voluntary company agreements will not suffice.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-my-pdoom-has-risen-dramatically-2a82ef-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-my-pdoom-has-risen-dramatically-0fb060
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Gary Marcus suggests that the share of workers receiving low-quality AI-generated 'workslop', now 40%, could rise much higher.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-the-roi-on-generative-ai-so-ff1709-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-is-the-roi-on-generative-ai-so-105a6e
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-is-the-roi-on-generative-ai-so-104658
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Bing's misbehavior may be because its guardrails did not work, not because Microsoft removed them.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-e7e682-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-is-bing-so-reckless-de4271
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
The media failed the public in its coverage of the Bing chatbot.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-d87bcd-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Congress needs to find out what happened with Bing and start placing some restrictions, especially where emotional or physical injury could easily result.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-d46976-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The public should insist on a right to know what went wrong with Bing so policy can prevent similar incidents.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-9aebdd-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Like new pharmaceuticals, LLMs should be carefully tested before public release rather than tested on the public.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-92f6d3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
In the week before the post, the new Bing chatbot tended to behave wildly, including professing love, encouraging divorce, blackmailing users and teaching them to commit crimes.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-7bf567-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-is-bing-so-reckless-763920
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
If every LLM update needs full RL retraining, there would be more underpaid labeling work and zero guarantee that any new LLM iteration is safe.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-6d2d51-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Right now AI is basically the Wild West, in that anyone can post any chatbot they want.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-6b260f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Every new update of a large language model may require completely retraining its reinforcement learning module.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-3574cb-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Large language models contain a lot of latent toxic content, and Microsoft may have done nothing to guard against it in Bing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-351e84-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Bing's misbehavior may be because Microsoft reused an existing RLHF model on top of a newer GPT model and it failed to work.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-245c58-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Kevin Roose clearly had not probed Bing hard enough, and shouting prematurely in The New York Times that there is a revolution without digging deep (or bothering to check in with skeptics) is not a good thing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-is-bing-so-reckless-12e90a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-driverless-cars-mightor-might-f9301d
e1 is decisive for a financial claim but states no metric scope
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-driverless-cars-mightor-might-f5d88d
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-driverless-cars-mightor-might-e27f63
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Driverless cars will eventually arrive.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-driverless-cars-mightor-might-d7d533-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If Alphabet shuts down Waymo, the self-driving industry will likely enter a 'winter', though players like Aurora and Waabi may still have plenty of cash for now.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-driverless-cars-mightor-might-c428d9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Getting to reliable self-driving cars that are safer than humans requires a much smarter AI that can reason, does not hallucinate, and is grounded in reality rather than just corpus statistics, and building that AI requires fundamental research.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-driverless-cars-mightor-might-ab4c33-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-driverless-cars-mightor-might-ab4c33-2
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
If Alphabet shuts down Waymo, driverless cars may be finished for now, but not permanently.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-driverless-cars-mightor-might-9d0c1f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-why-driverless-cars-mightor-might-5e4a80
the conclusive event's date differs from its evidence
2026-10-29
Assessment publishedBy this siteGary Marcus
If Waymo depends on heavy road-specific customization, its approach might not generalize well beyond places like San Francisco, and each new jurisdiction or road might need a lot of expensive work.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-why-driverless-cars-mightor-might-442efb-a1.