“Virtually none of the critical problems for AI that cognitive scientists such as Steve Pinker, Judea Pearl, the late Jerry Fodor, and myself have been pointing to for decades is actually solved yet.”
Our reading, the sentence that was checked:
Virtually none of the critical AI problems that cognitive scientists such as Steve Pinker, Judea Pearl, Jerry Fodor and Marcus have been pointing to for decades is actually solved yet.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The New Science of Alt Intelligence, paragraph 57.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-new-science-of-alt-intelligence-0f409d-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 2022-05-14. 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
- virtually none (scope, stated by the speaker)
- yet (qualifier, stated by the speaker)
From the speaker
- Virtually none
- critical problems for AI
- Steve Pinker, Judea Pearl, the late Jerry Fodor, and myself
- pointing to for decades
- actually solved yet
Our interpretation
- The immediate context identifies general text comprehension, modeling speaker intent, general reasoning and compositionality as examples of the problems.
- Solved means reliable competence across the relevant domain, rather than success on selected demonstrations or a narrow benchmark.
- Those examples guide the reading but cannot silently replace the larger, unspecified collection attributed to the named researchers.
- The target is May 2022 up to publication, the latest partial calendar month for this ongoing state. Older research is context.
Counts as supported if: A defensible inventory of the referenced problems and contemporary assessments establish that nearly all remained unsolved under their original scope.
Counts as contradicted if: Contemporary evidence establishes reliable solutions to a substantial portion of that inventory.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
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 horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The research supports caution about equating benchmark performance with solved general capabilities, while also documenting progress. It does not establish which problems belong in Marcus's collection or how many were solved. Restricting the claim to the next sentence's examples would make it easier to assess but would narrow his actual wording.
Limitations
- A complete inventory cannot be inferred from the researchers' names alone.
- Neither isolated successes nor isolated failures settle the proportion claimed.
Evidence (2 items)
Context only: Cambridge University Press,
casts doubts
Why it matters here: The survey explains why strong benchmark results need not establish robust language understanding. It supports caution about declaring problems solved, but does not count the problems Marcus references.
- Publisher
- Cambridge University Press
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: DeepMind / arXiv,
promise
Why it matters here: The authors report progress toward learning beyond trained tasks. This qualifies an overly sweeping reading of Marcus, but promising transfer is not a demonstrated solution to the broader problems.
- Publisher
- DeepMind / arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
An inventory grounded in the named researchers' original problem statements, followed by a problem-by-problem assessment of systems existing by May 14, 2022.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The New Science of Alt Intelligence
Virtually none of the critical problems for AI that cognitive scientists such as Steve Pinker, Judea Pearl, the late Jerry Fodor, and myself have been pointing to for decades is actually solved yet.
- Where
- paragraph 57
- Published
- 2022-05-14
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-new-science-of-alt-intelligence-0f409d-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-new-science-of-alt-intelligence-0f409d-a1.