“that LLM scaling would not bring us to AGI”
Our reading, the sentence that was checked:
Scaling LLMs will not produce AGI.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The bad news: AI is going pretty much as I expected, paragraph 2.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-bad-news-ai-is-going-pretty-much-5e7a47-a1
We recorded 7 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-09-05. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Forecast
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No deadline stated
- Deadline state on 2026-10-29
- No deadline stated (the release’s reference date)
From the speaker
- LLM scaling
- would not bring us to AGI
Our interpretation
- The plain reading is that increasing LLM scale will not by itself produce broadly human-level general intelligence.
- A decisive counterexample requires independent evidence of broad general capability and sufficient technical disclosure to attribute its achievement to LLM scaling.
- No deadline is supplied. Continued failures, expert skepticism, or achieving AGI through another method cannot establish this universal negative.
Counts as supported if: A broadly accepted impossibility result applicable to the intended class of scaled LLMs and to human-level general intelligence.
Counts as contradicted if: An independently evaluated system broadly recognized as human-level AGI, achieved through LLM scaling within the intended scope.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Did the forecast come true?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No deadline stated
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The search found major capability gains and public claims that AGI had arrived. The independent benchmark evaluator explicitly declined to call its result AGI, and the AGI Society found the published evidence insufficient. These records do not establish a decisive counterexample, but neither do they establish that scaling can never work. The forecast has no deadline.
Limitations
- There is no universally accepted AGI test.
- Public technical disclosures do not isolate the contribution of scaling.
- Failure to establish AGI by the evaluation date cannot support an unlimited negative forecast.
Evidence (3 items)
Context only: ARC Prize Foundation,
meaningful progress towards generalization
Why it matters here: The independent evaluation reports major progress, including 62.7% with its standard harness and 99.9% with a provider adapter, but explicitly rejects treating the result as proof of AGI. This is material progress without a decisive counterexample.
- Publisher
- ARC Prize Foundation
- Kind of source
- Benchmark result
- Describes events of
- Public from
Context only: AGI Society,
published evidence
Why it matters here: The assessment examines public claims that AGI had arrived and finds the available demonstrations insufficient. It documents a disputed achievement claim, not proof that scaling can never succeed.
- Publisher
- AGI Society
- Kind of source
- Secondary summary
- Describes events of
- Public from
Context only: OpenAI,
reinforcement learning
Why it matters here: The developer attributes the model to several research areas. The disclosure does not isolate increasing LLM scale as the sufficient cause of its capabilities; availability is conservatively dated to the page's September update.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
A dated, independently validated AGI demonstration with technical evidence that LLM scaling produced it, or an accepted impossibility result covering the specified route. Clarifying the meaning of scaling would reduce scope disputes.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The bad news: AI is going pretty much as I expected
that LLM scaling would not bring us to AGI
- Where
- paragraph 2
- Published
- 2025-09-05
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-bad-news-ai-is-going-pretty-much-5e7a47-a1: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.
About assessment marcus-the-bad-news-ai-is-going-pretty-much-5e7a47-a1.