“Melanie Mitchell’s book, Bender et al’s “Stochastic Parrots”; all of these more recent criticisms of deep learning help establish that some of the basic issues that were known long ago, with predecessors to current models, still remain even today.”
Our reading, the sentence that was checked:
Recent criticisms of deep learning, such as Melanie Mitchell's book and Bender et al.'s 'Stochastic Parrots', help establish that some of the basic issues known long ago with predecessors to current models still remain today.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Happy Groundhog Day, The AI Edition, paragraph 10.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-happy-groundhog-day-the-ai-edition-58c058-a1
We recorded 8 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 2023-02-03. 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
- some of the basic issues (scope, stated by the speaker)
- help establish (qualifier, stated by the speaker)
From the speaker
- Melanie Mitchell’s book
- Bender et al’s “Stochastic Parrots”
- help establish
- some of the basic issues
- still remain even today
Our interpretation
- The ordinary reading is modest: the named criticisms contribute evidence that at least some earlier limitations persist. It does not require every old limitation to remain or every model to fail.
- Relevant limitations include unreliable transfer, compositional generalization, and grounding of language.
- For the ongoing state, use 2022, the latest completed research year before the post and the year of the immediately preceding empirical example. Older criticisms are historical context. This coarse reporting-period choice is an interpretation because research has no fixed reporting cycle.
- The assessment includes the attribution to the named works, rather than merely checking whether any contemporary model had any weakness.
Counts as supported if: Verify relevant arguments or findings in the named works, match them to independently documented earlier limitations, and establish persistence in the contemporary research period.
Counts as contradicted if: The named works materially fail to support the proposed historical continuity, or matched contemporary evidence establishes that the identified limitations had been resolved.
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 retrieved research is consistent with the modest claim that some generalization problems persisted, and it also records neural-network successes. However, the full attribution to the named criticisms remains unverified because the relevant passages in Mitchell's book were not recovered. I would hold this assessment rather than treat a publisher description or another paper as a substitute for that evidence.
Limitations
- This is an evidence-access and attribution gap, not evidence that the historical continuity claim is false.
- The contemporary research example does not support a blanket claim of neural-network incapacity.
- The selected research year is an interpretation of the ongoing-state wording.
Evidence (2 items)
Context only: Cognition; author manuscript hosted by Rutgers University,
combinatorial structure
Why it matters here: The paper establishes that systematicity and structured representations were subjects of the earlier connectionist debate. It does not establish the status of models in 2022.
- Where
- Abstract
- Publisher
- Cognition; author manuscript hosted by Rutgers University
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: Proceedings of the National Academy of Sciences; abstract indexed by PubMed,
Hebbian
Why it matters here: The abstract reports compositional-generalization difficulties for standard networks and a neural model that captures benefits of human training curricula. It supports a qualified persistence claim while limiting any suggestion of universal failure, but does not verify the attribution to both named criticisms.
- Publisher
- Proceedings of the National Academy of Sciences; abstract indexed by PubMed
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
What would settle it
Obtain the relevant book passages and a clean comparison of both named works with historical sources and contemporary empirical findings about the same limitation.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Happy Groundhog Day, The AI Edition
Melanie Mitchell’s book, Bender et al’s “Stochastic Parrots”; all of these more recent criticisms of deep learning help establish that some of the basic issues that were known long ago, with predecessors to current models, still remain even today.
- Where
- paragraph 10
- Published
- 2023-02-03
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-happy-groundhog-day-the-ai-edition-58c058-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-happy-groundhog-day-the-ai-edition-58c058-a1.