“To be fair, large language models aren’t just looking things up, but the idea that a good prediction of next word necessarily implies understanding is fallacious.”
Our reading, the sentence that was checked:
Good next-word prediction by LLMs does not necessarily imply understanding.
Assessment: Can’t be checked Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, What was 60 Minutes thinking, in that interview with Geoff Hinton?, paragraph 81.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-what-was-60-minutes-thinking-in-that-40e9ec-a1
We recorded 20 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2023-10-10. 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)
From the speaker
- large language models aren’t just looking things up
- a good prediction of next word
- necessarily implies understanding
- is fallacious
Our interpretation
- Marcus disputes a necessary implication between prediction performance and understanding. He does not claim here that every LLM lacks every form of understanding.
- An ordinary reader can distinguish prediction success from comprehension, but the statement supplies neither a prediction threshold nor a sufficiently determinate meaning of understanding for a binary empirical verdict.
- Equating understanding with flawless performance on a selected benchmark would add a decisive premise Marcus did not supply. Intelligent humans can also make comprehension errors.
This claim is not graded on evidence. The record checks that the words are quoted correctly and attributed to the right person, and that the claim is classified correctly.
Current assessment
Not graded on evidence
Outcome: Can’t be checked
The statement is a recommendation, a value judgment, or too open to test. It is recorded, not scored.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
This is a claim about what prediction success logically establishes. The proposed empirical test would require an added definition of understanding. Repeated mistakes do not by themselves prove its absence, just as successful predictions do not settle its presence. A binary verdict would depend on the evaluator's philosophical or operational definition.
Limitations
- This does not reject Marcus's argument or endorse Hinton's.
- A narrower claim about a specified prediction measure and comprehension test could be assessed empirically.
Evidence (1 item)
No outcome evidence: a claim of this kind is recorded and attributed, not scored.
What would settle it
Specify the prediction threshold, the domain, and an independently measurable meaning of understanding, then assess the resulting narrower proposition.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, What was 60 Minutes thinking, in that interview with Geoff Hinton?
To be fair, large language models aren’t just looking things up, but the idea that a good prediction of next word necessarily implies understanding is fallacious.
- Where
- paragraph 81
- Published
- 2023-10-10
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-what-was-60-minutes-thinking-in-that-40e9ec-a1: Current. Outcome: Can’t be checked. 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-what-was-60-minutes-thinking-in-that-40e9ec-a1.