“But the problem is not going away; OpenAI’s recent o3 actually hallucinates more than some its predecesssors.”
Our reading, the sentence that was checked:
The hallucination problem is not going away; OpenAI's o3 hallucinates more than some of its predecessors.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Why DO large language models hallucinate?, paragraph 20.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-why-do-large-language-models-hallucinate-17bf2e-a1
We recorded 16 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 2025-05-05. 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
- the problem is not going away
- OpenAI’s recent o3
- hallucinates more than some its predecesssors
Our interpretation
- Read 'not going away' as persistence at the time of the post, rather than a separate prediction that hallucinations can never be reduced.
- Read 'more' as a higher measured hallucination rate on comparable factual-question evaluations against at least one earlier OpenAI model.
- Use the April 2025 release reporting period through the post date for the ongoing observation about 'recent o3'.
Counts as supported if: Comparable evaluations available for the release show o3 with a higher hallucination rate than an earlier OpenAI model.
Counts as contradicted if: Comparable release evaluations show no higher hallucination rate against the relevant predecessors, or corrected results invalidate the reported increases.
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
OpenAI's release evaluation reports higher hallucination rates for o3 than o1 on both PersonQA and SimpleQA. This supports the comparison and the observation that hallucinations persisted. The report also finds more correct answers from o3; the claim should not be extended to mean that o3 was less accurate on every measure or that future progress was impossible.
Limitations
- These are provider-reported results on two factual-question benchmarks.
- The report does not establish a universal hallucination rate across tasks or a rate per individual assertion.
- The event date identifies the evaluated release; individual test execution dates are not supplied.
Evidence (1 item)
Supports the claim: OpenAI,
Specifically, o3 tends to make more claims overall, leading to more accurate claims as well as more inaccurate/hallucinated claims.
Why it matters here: The release report gives PersonQA hallucination rates of 0.33 for o3 and 0.16 for o1, and SimpleQA rates of 0.51 and 0.44. It also reports higher accuracy for o3 on both tests, so increased hallucination does not mean uniformly worse performance.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Why DO large language models hallucinate?
But the problem is not going away; OpenAI’s recent o3 actually hallucinates more than some its predecesssors.
- Where
- paragraph 20
- Published
- 2025-05-05
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-why-do-large-language-models-hallucinate-17bf2e-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-why-do-large-language-models-hallucinate-17bf2e-a1.