“All of this – the hallucinations, mistakes, and overconfidence – is entirely typical of LLMs, and entirely problematic in medicine.”
Our reading, the sentence that was checked:
Hallucinations, mistakes and overconfidence are entirely typical of LLMs and entirely problematic in medicine.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Please don’t trust your chatbot for medical advice, paragraph 6.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-please-dont-trust-your-chatbot-for-7ffa80-a1
We recorded 6 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 2026-04-21. 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)
From the speaker
- hallucinations, mistakes, and overconfidence
- entirely typical of LLMs
- entirely problematic in medicine
Our interpretation
- Typical means characteristic, recurring failure modes across models, not behavior present in every answer or a claim that most answers are false.
- Problematic in medicine means capable of undermining reliable medical information or decisions; it does not require demonstrated patient injury.
- The contextual reference to approximately year-old testing identifies a 2025-to-publication evidence period. Generalization to all current versions remains qualified.
Counts as supported if: Independent evaluations across multiple LLMs document recurring fabricated content, factual errors and unjustified certainty in medically relevant tasks.
Counts as contradicted if: Broad comparable evaluations show these failures are exceptional, confined to isolated systems, or immaterial to medical information and decisions.
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
The evidence supports recurring failure modes across models, with clear relevance to medical reliability. It supports characteristic risks, not the stronger readings that every model always exhibits them or that most medical answers are wrong.
Limitations
- The decisive audit uses misinformation-oriented prompts and older model versions.
- Disclaimers and stronger performance on some topics qualify the categorical wording.
- The independent corroborating study is a preprint and measures medical evidence synthesis.
Evidence (2 items)
Supports the claim: BMJ Open,
misinformation.
Why it matters here: The audit documents all three failure types across popular chatbots, including fabricated citations and certainty accompanying incorrect answers. Professional-consultation disclaimers nevertheless appeared frequently, qualifying any literal claim of universal certainty.
- Publisher
- BMJ Open
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: arXiv,
uncertain effect.
Why it matters here: The publicly released 24-model evaluation independently reports reluctance to express uncertainty when synthesizing medical evidence. The date identifies the report's release; exact model-query dates were not established, so it is corroborating context rather than decisive dated outcome evidence.
- Publisher
- arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Please don’t trust your chatbot for medical advice
All of this – the hallucinations, mistakes, and overconfidence – is entirely typical of LLMs, and entirely problematic in medicine.
- Where
- paragraph 6
- Published
- 2026-04-21
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-please-dont-trust-your-chatbot-for-7ffa80-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-please-dont-trust-your-chatbot-for-7ffa80-a1.