“There is no internal model of what she actually did, and hence no capacity to support reliable inference, hence no guarantee that anything it writes about her will be true.”
Our reading, the sentence that was checked:
LinkedIn's OpenAI-powered assistant has no internal model of what Kaya Yurieff actually did, and hence no capacity to support reliable inference about her, and hence no guarantee that anything it writes about her will be true.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Muddles about Models, paragraph 26.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-muddles-about-models-244cf7-a2
We recorded 8 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 2023-10-05. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Causal claim
- 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
- about Kaya Yurieff (scope, stated by the speaker)
From the speaker
- no internal model of what she actually did
- hence no capacity to support reliable inference
- hence no guarantee that anything it writes about her will be true
Our interpretation
- The claim concerns the LinkedIn assistant in the test discussed immediately before the October 2023 post, specifically its representations and inferences about Kaya Yurieff.
- The plain reading is that the absence of a valid representation of her history causes unreliable biographical inference.
- No guarantee means individual statements cannot be assumed accurate; it does not mean every statement must be false.
- The supplied context does not establish the exact test date. Later product behavior cannot settle the historical causal claim.
Counts as supported if: Evidence from the tested assistant establishes the absence of a usable representation of Yurieff’s history and links that absence to unreliable inferences about her.
Counts as contradicted if: Technical examination of that assistant establishes a usable representation of her history and controlled tests show reliable inference from it, undermining the claimed causal chain.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Does the evidence support the cause?
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
LinkedIn’s March 15, 2023 announcement describes writing suggestions based on existing profile content and advises users to review them for accuracy [e1]. It provides no examination of the assistant’s representations of Yurieff’s history or controlled tests of its inferences about her [e1]. This evidence does not establish the claimed causal chain [e1].
Limitations
- No evidence identifying the version, inputs, prompts or complete outputs of the assistant tested by Yurieff.
- No technical examination or controlled intervention establishing whether missing representations caused errors.
- The accuracy advice in e1 does not establish an absence of reliable inference capacity.
Evidence (1 item)
Context only: LinkedIn; Tomer Cohen,
writing suggestions.
Why it matters here: The announcement describes profile-based suggestions and advises users to check and edit them for accuracy. It supplies product context, but no examination of the assistant’s representations or inferences about Yurieff.
- Publisher
- LinkedIn; Tomer Cohen
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
The original test record, identified assistant version, profile inputs, prompts and verified biographical facts, together with technical examination and repeated controlled tests distinguishing missing information from failure to use available information.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Muddles about Models
There is no internal model of what she actually did, and hence no capacity to support reliable inference, hence no guarantee that anything it writes about her will be true.
- Where
- paragraph 26
- Published
- 2023-10-05
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-muddles-about-models-244cf7-a2: 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.