“The second example illustrates that it cannot reliably produce a non-standard number of body parts.”
Our reading, the sentence that was checked:
ChatGPT with DALL-E 3 cannot reliably produce images with a non-standard number of body parts.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Don’t Ride This Bike! Generative AI’s persistent trouble with compositionality and parts, paragraph 28.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-dont-ride-this-bike-generative-ais-ec6b8b-a2
We recorded 10 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 2024-12-08. 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
- based on a small number of examples (qualifier, stated by the speaker)
From the speaker
- cannot reliably produce a non-standard number of body parts
- Draw a picture of a rabbit with four ears
- on the basis of a small number of examples
Our interpretation
- It refers to ChatGPT with DALL-E 3.
- The plain reading is recurring failure when explicitly asked to depict an unusual anatomical count. It does not assert that every such request fails.
- The target period is November 2024, the month explicitly identified for the underlying experiments.
- Counting whole animals or accidentally generating malformed anatomy does not directly test compliance with a requested nonstandard body-part count.
Counts as supported if: Relevant November 2024 outputs show repeated failures to produce explicitly requested nonstandard body-part counts.
Counts as contradicted if: Comparable testing shows consistently correct requested nonstandard counts across repeated attempts and relevant anatomical variations.
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 study released on November 26, 2024 tests requested numbers of whole objects and reports frequent failures for numbers beyond three [e2]. It does not test explicitly requested nonstandard body-part counts or isolate the November ChatGPT deployment [e2]. Its results therefore do not settle the claimed anatomical-count reliability [e2].
Limitations
- No verified contemporary prompt-and-image evaluation directly measures requested nonstandard anatomical counts.
- The original rabbit and two-headed-deer images were not independently verified.
- Whole-object counting results do not establish body-part counting reliability.
Evidence (1 item)
Context only: arXiv; Colin Conwell, Rupert Tawiah-Quashie, and Tomer Ullman,
The negation probes and numbers (beyond 3) fail most frequently.
Why it matters here: The study reports difficulties with counting whole objects in experiments conducted during March–November 2024. It does not measure requested nonstandard body-part counts or isolate the November ChatGPT interface.
- Publisher
- arXiv; Colin Conwell, Rupert Tawiah-Quashie, and Tomer Ullman
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
Verified November 2024 prompt-and-image records from the relevant ChatGPT deployment, covering repeated requests for unusual body-part counts, preserving all attempts and including both successes and failures.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Don’t Ride This Bike! Generative AI’s persistent trouble with compositionality and parts
The second example illustrates that it cannot reliably produce a non-standard number of body parts.
- Where
- paragraph 28
- Published
- 2024-12-08
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-dont-ride-this-bike-generative-ais-ec6b8b-a2: 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-dont-ride-this-bike-generative-ais-ec6b8b-a2.