“the fact that LLMs hallucinate (patently true)”
Our reading, the sentence that was checked:
LLMs hallucinate.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Is the US military actually afraid of Claude? A new theory of why Anthropic was labeled a supply chain risk., paragraph 10.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-is-the-us-military-actually-afraid-6e3132-a1
We recorded 10 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2026-03-12. 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
- LLMs hallucinate
- patently true
Our interpretation
- Hallucination means generating fabricated or factually false information as an answer.
- The ordinary reading is that hallucinations occur in LLMs, without claiming that every model or every response hallucinates.
- Use the current calendar quarter through publication, January 1–March 12, 2026, as the reporting window for the ongoing technical state. No fixed reporting cycle is specified.
Counts as supported if: A documented evaluation or checked example shows hallucinations in publicly available LLMs during the target period.
Counts as contradicted if: Reliable evidence establishes that the alleged hallucinations are not factual fabrications or errors and that the relevant models do not exhibit the claimed behavior. A lower rate or a clean finite test is insufficient.
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 pre-post release evaluation explicitly reports hallucinations and partial reductions in their frequency. This establishes that the behavior occurs. Improvement does not contradict Marcus's claim, which states no minimum rate.
Limitations
- The evidence is a manufacturer's internal evaluation.
- The selected prompts emphasize previously reported factual failures and do not establish an everyday population-wide error rate.
- The result does not establish that every LLM hallucinates on every task.
Evidence (1 item)
Supports the claim: OpenAI,
hallucinations decrease by 22.5% with web use and 9.6% without web access.
Why it matters here: OpenAI reports hallucinations in its release evaluation, including after improvements. The reported reductions imply residual hallucinations; they are not claims of elimination.
- 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, Is the US military actually afraid of Claude? A new theory of why Anthropic was labeled a supply chain risk.
the fact that LLMs hallucinate (patently true)
- Where
- paragraph 10
- Published
- 2026-03-12
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-is-the-us-military-actually-afraid-6e3132-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-is-the-us-military-actually-afraid-6e3132-a1.