“Almost everything it says is fluent and at least vaguely plausible.”
Our reading, the sentence that was checked:
Almost all of GPT's output is fluent and at least vaguely plausible, not random or word salad.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, How come GPT can seem so brilliant one minute and so breathtakingly dumb the next?, paragraph 25.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-how-come-gpt-can-seem-so-brilliant-8af3b5-a1
We recorded 9 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 2022-12-01. 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
- Almost everything it says
- fluent
- at least vaguely plausible
Our interpretation
- The plain reading is a near-universal property of ordinary responses, with only rare exceptions, rather than merely an ability to produce good examples.
- Both fluency and superficial plausibility must hold. Incorrect statements can still be superficially plausible.
- Marcus gives no percentage, so I do not invent one.
- Use November 30 through December 1, 2022, the launch-period observation window described by the surrounding discussion.
Counts as supported if: A sufficiently broad and unselected sample of contemporary responses shows that failures of either fluency or superficial plausibility were rare.
Counts as contradicted if: A sufficiently broad and unselected sample shows a substantial recurring share of responses failing fluency or superficial plausibility, beyond rare exceptions.
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
Contemporary records establish fluent and plausible-sounding output, including fluent errors. They do not establish how close to universal that quality was. Selected successes cannot support almost everything, and isolated nonsensical answers cannot contradict a claim that allows rare exceptions.
Limitations
- The unresolved result reflects missing frequency evidence, not evidence that the claim was false.
- Factual-error rates would not directly measure this claim about surface quality.
Evidence (2 items)
Context only: OpenAI,
plausible-sounding but incorrect or nonsensical answers.
Why it matters here: The disclosure confirms that surface plausibility can coexist with errors. It does not measure how often responses satisfy both fluency and plausibility, and nonsensical content need not be word salad.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: The Verge,
still prone to producing cogent waffle and misinformation.
Why it matters here: The report provides contemporary observations of fluid responses alongside invented information. It is useful corroboration, but neither its examples nor its qualitative description establish the claimed near-universal rate.
- Publisher
- The Verge
- Kind of source
- News report
- Describes events of
- Public from
What would settle it
A broad, unselected archive of responses from the launch-period model, with a documented prompt population and independent ratings of both fluency and superficial plausibility.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, How come GPT can seem so brilliant one minute and so breathtakingly dumb the next?
Almost everything it says is fluent and at least vaguely plausible.
- Where
- paragraph 25
- Published
- 2022-12-01
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-how-come-gpt-can-seem-so-brilliant-8af3b5-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-how-come-gpt-can-seem-so-brilliant-8af3b5-a1.