“The core of the paper challenged the idea that LLMs can reason on any deep level, and showed that they broke down on moderately complex versions of basic algorithms (since Tower of Hanoi with 8 discs, first solved with classical techniques around 1957) — casting serious doubt into whether LLMs on their own could ever achieve AGI.”
Our reading, the sentence that was checked:
The Apple paper challenged the idea that LLMs can reason on any deep level and showed that they broke down on moderately complex versions of basic algorithms such as Tower of Hanoi with 8 discs, which casts serious doubt on whether LLMs on their own could ever achieve AGI.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Five quick updates about that Apple reasoning paper that people can’t stop talking about, paragraph 2.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-five-quick-updates-about-that-apple-0e24e7-a1
We recorded 9 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 2025-06-17. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
Conditions
- as shown by the Apple paper (scope, stated by the speaker)
- moderately complex versions of basic algorithms (e.g., Tower of Hanoi with 8 discs) (scope, stated by the speaker)
- LLMs on their own (qualifier, stated by the speaker)
- casting serious doubt (not asserting impossibility) (qualifier, stated by the speaker)
From the speaker
- The core of the paper challenged
- broke down on moderately complex versions of basic algorithms
- Tower of Hanoi with 8 discs
- LLMs on their own
- casting serious doubt
Our interpretation
- Assess the findings reported in the Apple paper discussed in June 2025, as available by the post date.
- The plain reading is that the experiments exposed substantial reasoning limitations in the tested models. Casting doubt on AGI is an inference, not a forecast that AGI is impossible.
- Moderately complex means relatively small instances of familiar algorithmic puzzles, including approximately eight Hanoi discs; it does not mean the paper's formally named medium-complexity regime.
- The parenthetical reference to 1957 is ancillary historical background and is not independently resolved by this assessment.
Counts as supported if: The paper reports substantial failure on the specified puzzle tasks around the stated complexity, including failures in executing supplied algorithms.
Counts as contradicted if: The paper reports reliable success at that complexity, or does not report the algorithmic failures attributed to it.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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 horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The empirical description matches the original paper: it reports collapse around the cited Hanoi complexity and failures even with a supplied algorithm. That supports a challenge to the tested models' reasoning reliability. It does not establish that all deep reasoning is absent or that future LLM-only AGI is impossible.
Limitations
- Support applies to the reported experiments and their challenge to reasoning reliability.
- Alternative output formats test different capabilities and weaken sweeping conclusions.
- The ancillary historical date is unverified.
Evidence (2 items)
Supports the claim: Apple researchers, arXiv,
Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities.
Why it matters here: Figure 7 reports near-zero intermediate-solution accuracy for Hanoi at eight or more discs. Section 4.4 reports that supplying the algorithm did not remove the collapse.
- Publisher
- Apple researchers, arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: C. Opus and A. Lawsen, arXiv,
Full experimental validation remains as future work.
Why it matters here: The rejoinder reports preliminary success when models generate a Hanoi-solving function instead of listing moves. This challenges generalization from Apple's task to reasoning overall, while acknowledging limited experimental validation.
- Publisher
- C. Opus and A. Lawsen, arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Five quick updates about that Apple reasoning paper that people can’t stop talking about
The core of the paper challenged the idea that LLMs can reason on any deep level, and showed that they broke down on moderately complex versions of basic algorithms (since Tower of Hanoi with 8 discs, first solved with classical techniques around 1957) — casting serious doubt into whether LLMs on their own could ever achieve AGI.
- Where
- paragraph 2
- Published
- 2025-06-17
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-five-quick-updates-about-that-apple-0e24e7-a1: Current. Outcome: Supported. 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-quick-updates-about-that-apple-0e24e7-a1.