“Even “draw a white bicycle with green wheels”, which focuses purely on part-whole relationships without function or complex syntax, is problematic:”
Our reading, the sentence that was checked:
Stable Diffusion has trouble even with simple part-whole prompts such as a white bicycle with green wheels.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Form, function, and the giant gulf between drawing a picture and understanding the world, paragraph 35.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-form-function-and-the-giant-gulf-a558ad-a1
We recorded 6 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 2022-09-18. 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
- draw a white bicycle with green wheels
- part-whole relationships
- without function or complex syntax
- is problematic
Our interpretation
- The ordinary reading is that contemporary Stable Diffusion has recurring difficulty assigning requested colors separately to an object and its parts.
- For the named example, successful output should depict a white bicycle body or frame with recognizably green wheels.
- The claim does not require universal failure, and a single successful image would not refute it.
- The post describes ongoing September 2022 behavior without a reporting period. I use September through the publication date as the observation window.
- Difficulty assigning colors to separate objects is relevant background but does not by itself establish difficulty assigning colors to an object and its parts.
Counts as supported if: Independent contemporary outputs show repeated failures on the named bicycle prompt or closely comparable simple object-and-part color prompts during the target period.
Counts as contradicted if: Representative independent contemporary testing shows reliable assignment of the requested colors to the bicycle and its wheels, together with comparable simple object-and-part prompts.
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 available research concerns a related color-assignment problem, but it does not directly test the claimed part-whole behavior during the target period. Searches for the exact example did not locate an independent contemporary reproduction. Neither support nor contradiction is established by the qualifying records.
Limitations
- Failure to locate independent records does not show that Marcus's displayed results were inaccurate.
- Multiple-object color binding and object-part color assignment are related but distinct tests.
- The original model settings and complete sample are unavailable in the supplied text.
Evidence (1 item)
Context only: arXiv; T2I-CompBench authors,
confuse the colors
Why it matters here: This research identifies a related color-assignment problem. Its later experiments and multiple-object task do not directly establish September 2022 performance on bicycle-and-wheel color assignment.
- Publisher
- arXiv; T2I-CompBench authors
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
Independent September 1–18, 2022 outputs with model and sampling details for the white-bicycle/green-wheels prompt or closely comparable object-and-part color prompts, showing whether failures recur.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Form, function, and the giant gulf between drawing a picture and understanding the world
Even “draw a white bicycle with green wheels”, which focuses purely on part-whole relationships without function or complex syntax, is problematic:
- Where
- paragraph 35
- Published
- 2022-09-18
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-form-function-and-the-giant-gulf-a558ad-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-form-function-and-the-giant-gulf-a558ad-a1.