“There is no *principled* solution thus far to any of them.”
Our reading, the sentence that was checked:
None of the known problems discussed in Marcus's thread has a principled solution thus far.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Why I don’t share Sam Altman’s confidence that AGI is basically a solved problem, paragraph 7.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-sam-altman-thinks-that-agi-is-basically-97196d-a1
We recorded 16 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 2025-01-06. 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
- the problems discussed in the thread (scope, stated by the speaker)
- principled (qualifier, stated by the speaker)
- thus far (qualifier, stated by the speaker)
From the speaker
- no *principled* solution
- thus far
- to any of them
Our interpretation
- The ordinary reading is that none of the discussed problems had a demonstrated, systematic solution by publication. It does not mean that no mitigation or promising research existed.
- The scope includes distribution shift, commonsense reasoning, reasoning-model robustness, generalization beyond domains supporting data augmentation, scaling limitations, and hallucinations. The thread also mixes in economic and presentation concerns.
- For the ongoing technical state, the target is the 2024 research cycle through publication on January 6, 2025. Earlier methods remain relevant background.
- A principled solution means a method with an explanatory basis and evidence of reliable performance across the relevant problem class. Marcus supplies neither an acceptance threshold nor a precise boundary for each problem.
Counts as supported if: A sufficiently comprehensive record establishes that every identified problem remained unresolved under the stated reading, after considering proposed systematic solutions.
Counts as contradicted if: At least one identified problem had a demonstrated systematic solution with sufficient scope and reliability by January 6, 2025.
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 evidence establishes active mitigation research and restricted successes, not a general solution to the listed problems. It also cannot establish Marcus's universal negative. Calling every existing method a workaround would require an acceptance standard he did not provide. The broad claim therefore remains unresolved.
Limitations
- This finding does not dispute the narrower observations of continuing reliability failures.
- The result depends on distinguishing a principled approach from a demonstrated solution.
Evidence (2 items)
Context only: arXiv; Tonmoy and colleagues,
over 32 techniques
Why it matters here: The survey documents numerous systematic mitigation methods and discusses their limitations. Their existence challenges a reading of no principled approaches, but does not establish a complete solution to hallucinations.
- Where
- Abstract
- Publisher
- arXiv; Tonmoy and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: Nature; Brenden Lake and Marco Baroni,
several systematic generalization benchmarks.
Why it matters here: The authors demonstrate systematic generalization through meta-learning on specified tasks. This is substantive prior progress, but neither its scope nor its date makes it decisive evidence of a general solution for contemporary LLMs.
- Where
- Abstract
- Publisher
- Nature; Brenden Lake and Marco Baroni
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
What would settle it
A defined inventory of the problems and a systematic review of pre-publication methods against explicit scope and reliability criteria.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Why I don’t share Sam Altman’s confidence that AGI is basically a solved problem
There is no *principled* solution thus far to any of them.
- Where
- paragraph 7
- Published
- 2025-01-06
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-sam-altman-thinks-that-agi-is-basically-97196d-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-sam-altman-thinks-that-agi-is-basically-97196d-a1.