“A key claim of the paper was that pure scaling of LLMs – just adding more data and compute to extant LLM architectures – would not bring us to AGI, and that so-called scaling laws were merely empirical generalizations rather than physical laws.”
Our reading, the sentence that was checked:
Pure scaling of LLMs by adding data and compute to existing architectures will not reach AGI; scaling laws are empirical generalizations, not physical laws.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Five ways in which the last 3 months — and especially the DeepSeek era — have vindicated “Deep learning is hitting a wall", paragraph 4.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-five-ways-in-which-the-last-3-months-bc7c31-a1
We recorded 8 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 2025-02-09. 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)
Conditions
- pure scaling with extant architectures (qualifier, stated by the speaker)
From the speaker
- pure scaling of LLMs
- just adding more data and compute to extant LLM architectures
- would not bring us to AGI
- merely empirical generalizations rather than physical laws
Our interpretation
- The forecast concerns increasing training data and compute without additional reasoning methods or other substantive changes.
- AGI means broadly general, human-level intellectual capability established through independent evaluation, rather than success on one benchmark.
- The empirical status of scaling laws is a separate, presently checkable clause. Establishing it does not establish the forecast.
- Pure scaling restricts the proposed route to AGI; it is not a separate contingent antecedent.
Counts as supported if: A sufficiently general result establishing that the specified scaling-only approach cannot attain AGI would support the forecast. Research describing fitted empirical relationships supports only the scaling-law clause.
Counts as contradicted if: A system independently established as AGI, with documentation showing that its qualifying capabilities arose solely from increased data and training compute on the relevant existing LLM architectures.
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 research supports the description of scaling laws as empirical relationships. It does not establish that scaling cannot reach AGI. No decisive event meeting the resolution rule was established by the evaluation date. An indefinite negative forecast cannot be credited merely because a qualifying success has not been demonstrated.
Limitations
- Public model documentation may not disclose enough to establish scaling-only provenance.
- The empirical clause and the AGI forecast have different evidentiary requirements.
Evidence (1 item)
Context only: Kaplan et al., arXiv,
empirical scaling laws
Why it matters here: The researchers explicitly describe empirical relationships for language-model loss. This supports that clause but supplies no impossibility result about reaching AGI through scaling.
- Publisher
- Kaplan et al., arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
Independent validation of AGI accompanied by a documented scaling-only development history, or a general impossibility result covering the specified approach.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Five ways in which the last 3 months — and especially the DeepSeek era — have vindicated “Deep learning is hitting a wall"
A key claim of the paper was that pure scaling of LLMs – just adding more data and compute to extant LLM architectures – would not bring us to AGI, and that so-called scaling laws were merely empirical generalizations rather than physical laws.
- Where
- paragraph 4
- Published
- 2025-02-09
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-five-ways-in-which-the-last-3-months-bc7c31-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-five-ways-in-which-the-last-3-months-bc7c31-a1.