“There is no evidence that the problem was heretofore “unsolvable””
Our reading, the sentence that was checked:
There is no evidence that the problem FunSearch addressed was previously unsolvable.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Sorry, but FunSearch probably isn’t a milestone in scientific discovery, paragraph 12.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-sorry-but-funsearch-probably-isnt-2de8a6-a1
We recorded 7 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2023-12-19. 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
- There is no evidence
- the problem
- heretofore “unsolvable”
Our interpretation
- The problem is the cap-set construction and lower-bound problem discussed in the FunSearch report.
- The plain reading distinguishes impossible to solve from previously unsolved or lacking a particular record.
- Heretofore refers to the period before FunSearch's December 14, 2023 publication; the historical interval has no specified starting date.
- The absence claim is assessed against relevant published mathematical work, rather than every possible unpublished statement.
Counts as supported if: Relevant prior mathematical work documents constructive progress on cap-set lower bounds, and the searched record provides no substantiated impossibility result for the task FunSearch addressed.
Counts as contradicted if: Relevant pre-FunSearch evidence establishes that the task was impossible under the same assumptions, or otherwise substantiates the stronger description unsolvable.
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 relevant literature documents earlier constructive improvements, and the search found no substantiated impossibility result. That supports Marcus's distinction between unsolvable and previously unsolved. FunSearch nevertheless produced a new improvement. If unsolvable is read loosely as a particular result not previously achieved, the assessment would change.
Limitations
- Support is limited to the ordinary distinction between impossibility and an open problem.
- Prior improvements alone do not prove that every later target bound was attainable.
- An absence claim cannot be verified exhaustively.
Evidence (2 items)
Supports the claim: arXiv / Fred Tyrrell,
new lower bound
Why it matters here: This predating result demonstrates constructive progress on the same lower-bound problem. It supports treating the task as an open optimization question rather than an established impossibility.
- Publisher
- arXiv / Fred Tyrrell
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: Discrete Analysis,
improved the constant
Why it matters here: The journal identifies a real new improvement by FunSearch. This limits the assessment: Marcus's distinction about impossibility does not establish that the particular new bound was already known.
- Publisher
- Discrete Analysis
- Kind of source
- Secondary summary
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Sorry, but FunSearch probably isn’t a milestone in scientific discovery
There is no evidence that the problem was heretofore “unsolvable”
- Where
- paragraph 12
- Published
- 2023-12-19
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-sorry-but-funsearch-probably-isnt-2de8a6-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-sorry-but-funsearch-probably-isnt-2de8a6-a1.