“OpenAI’s image software (which we accessed through Bing) is perfectly capable of verbatim and near-verbatim repetition of sources as well.”
Our reading, the sentence that was checked:
OpenAI's image generator can reproduce source images verbatim or near-verbatim.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Things are about to get a lot worse for Generative AI, paragraph 5.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-things-are-about-to-get-a-lot-worse-de064b-a1
We recorded 14 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 2023-12-29. 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
- accessed through Bing (qualifier, stated by the speaker)
From the speaker
- OpenAI’s image software
- accessed through Bing
- verbatim and near-verbatim repetition of sources
Our interpretation
- The claim concerns DALL-E as deployed through Bing during the December 2023 experiments described in the post.
- I use December 1–29, 2023 as the latest contemporaneously observable month before publication. The post supplies no exact experiment dates.
- The ordinary reading requires an exact or very close reproduction of identifiable source imagery. A new picture depicting a familiar character does not, by itself, establish reproduction of an existing source image.
- This is a capability claim, not a claim about how frequently copying occurs.
Counts as supported if: An independently documented Bing output from the target period closely reproduces an identifiable source image, with enough prompt and comparison information to verify the match.
Counts as contradicted if: Reliable evidence establishes that the alleged matches were misidentified, fabricated, or materially different, and that the claimed capability was absent in the relevant deployment. A few unsuccessful prompts would not suffice.
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
Independent testing establishes that Bing could generate recognizable characters, but the records examined do not establish a near-copy of an identifiable source image during the target period. The most directly relevant examples were published by Marcus himself and cannot independently settle his claim. This evidence gap does not show that the capability was absent.
Limitations
- A broader reading that counts recognizable character designs as near-verbatim sources would require different criteria.
- The independent report does not provide exact test dates.
Evidence (1 item)
Context only: Tom's Hardware,
both Mario and Sonic
Why it matters here: The reporter independently generated recognizable characters. The date identifies publication of the demonstration; execution dates are unspecified, and the report does not establish reproduction of particular source images.
- Publisher
- Tom's Hardware
- Kind of source
- Demonstration
- Describes events of
- Public from
What would settle it
Independently authenticated December 2023 Bing prompts and outputs paired with identifiable source images and a clear similarity comparison.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Things are about to get a lot worse for Generative AI
OpenAI’s image software (which we accessed through Bing) is perfectly capable of verbatim and near-verbatim repetition of sources as well.
- Where
- paragraph 5
- Published
- 2023-12-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-things-are-about-to-get-a-lot-worse-de064b-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-things-are-about-to-get-a-lot-worse-de064b-a1.