“As is by now widely known, they are a hallmark of systems that generalize based on word distribution.”
Our reading, the sentence that was checked:
As is by now widely known, hallucinations are a hallmark of systems that generalize based on word distribution.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Hello, Multimodal Hallucinations, paragraph 10.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-hello-multimodal-hallucinations-5f7eea-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 2023-10-21. 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
- as is by now widely known (qualifier, stated by the speaker)
From the speaker
- As is by now widely known
- hallmark
- systems that generalize based on word distribution
Our interpretation
- The ordinary reading is that hallucination was a recognized, recurring failure of contemporary language models trained to predict text.
- Hallmark does not require every answer or every model to hallucinate.
- The sentence describes an observed characteristic; it does not establish that word-distribution learning is the sole cause.
- For this ongoing research state, I use the current reporting year through publication, January 1–October 21, 2023. There is no standard quarterly reporting period for this literature.
Counts as supported if: Contemporaneous research and model evaluations identify hallucinations as a recurring, broadly recognized limitation across language models.
Counts as contradicted if: Representative contemporary evaluations establish that hallucination was exceptional rather than characteristic, and the claimed broad recognition is unsupported.
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
A September 2023 research survey treated hallucination as a substantial LLM reliability problem, while OpenAI's GPT-4 report directly acknowledged the same limitation. Together they support the ordinary reading of a widely recognized, recurring failure. Reported improvements do not negate that finding.
Limitations
- This establishes neither a single hallucination rate nor that all models perform equally poorly.
- It does not prove that distributional learning alone causes hallucination or makes it inevitable.
Evidence (2 items)
Supports the claim: arXiv; Yue Zhang and colleagues,
a substantial challenge
Why it matters here: The survey identifies hallucination as a substantial reliability problem and organizes existing detection, explanation and mitigation research. This supports broad recognition of a recurring failure, without supplying a universal frequency.
- Publisher
- arXiv; Yue Zhang and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Supports the claim: OpenAI,
“hallucinates” facts
Why it matters here: OpenAI explicitly acknowledges hallucination in GPT-4 and compares its limitations with earlier GPT models. The same section reports improvements over GPT-3.5, which qualify severity but do not contradict persistence.
- 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, Hello, Multimodal Hallucinations
As is by now widely known, they are a hallmark of systems that generalize based on word distribution.
- Where
- paragraph 10
- Published
- 2023-10-21
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-hello-multimodal-hallucinations-5f7eea-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-hello-multimodal-hallucinations-5f7eea-a1.