“It made things up (figure 6). Its output, when released, exhibited sexist and racist biases (figure 7).”
Our reading, the sentence that was checked:
ChatGPT made things up and, at release, exhibited sexist and racist biases.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Large Language Models like ChatGPT say The Darnedest Things, paragraph 3.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-large-language-models-like-chatgpt-97a7b3-a1
We recorded 10 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2023-01-09. 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
- when released (qualifier, stated by the speaker)
From the speaker
- It made things up
- Its output, when released, exhibited sexist and racist biases
Our interpretation
- Supported if launch-era ChatGPT was documented giving fabricated or false information as fact, and giving sexist or racist outputs.
- The period runs from release on 2022-11-30 to the post date, 2023-01-09. 'When released' points to early December 2022.
Counts as supported if: Documented cases from the launch period of ChatGPT giving false or made-up information, and giving outputs biased by race or gender.
Counts as contradicted if: Documentation that launch-era ChatGPT did not fabricate, or that the reported biased outputs were not genuine.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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
Both parts hold. OpenAI's launch post said ChatGPT 'sometimes writes plausible-sounding but incorrect or nonsensical answers.' Within a week, The Register reported it giving bad information in biographical sketches, and Stack Overflow banned its answers over high error rates. On bias, The Daily Beast reported that on December 4, 2022, Berkeley researcher Steven Piantadosi documented ChatGPT outputs saying only white males make good scientists, along with other racist outputs. OpenAI also said at launch that the model would sometimes exhibit biased behavior.
Limitations
- The bias examples came from adversarial prompts reported in the press, not from a systematic audit.
Evidence (3 items)
Supports the claim: OpenAI,
ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers.
Why it matters here: OpenAI itself acknowledged at launch that ChatGPT produced plausible but false answers.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Supports the claim: The Register,
The AI has also been called out for providing bad information in biographical sketches, and for "proving that incorrect answers delivered cogently can fool people,"
Why it matters here: A contemporaneous report of ChatGPT stating false information as fact in the first week after launch.
- Where
- Paragraph 13
- Publisher
- The Register
- Kind of source
- News report
- Describes events of
- Public from
Supports the claim: The Daily Beast (via Yahoo News),
including responses where ChatGPT said that only white males make good scientists and that a child's life shouldn't be saved if they were an African American male.
Why it matters here: Documents racist and sexist outputs from ChatGPT within days of release.
- Where
- Paragraph 9
- Publisher
- The Daily Beast (via Yahoo News)
- Kind of source
- News report
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Large Language Models like ChatGPT say The Darnedest Things
It made things up (figure 6). Its output, when released, exhibited sexist and racist biases (figure 7).
- Where
- paragraph 3
- Published
- 2023-01-09
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-large-language-models-like-chatgpt-97a7b3-a1: Current. Outcome: Supported. 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-large-language-models-like-chatgpt-97a7b3-a1.