““new effort by OpenAI to solve these problems wound [still] fabricat[ing] authoritative nonsense ” Still true.”
Our reading, the sentence that was checked:
OpenAI's new effort to solve these problems still ends up fabricating authoritative nonsense.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Two years later, deep learning is still faced with the same fundamental challenges, paragraph 19.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-two-years-later-deep-learning-is-d4c2a5-a2
We recorded 15 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 2024-03-10. 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
- OpenAI's new effort to solve these problems (scope, stated by the speaker)
From the speaker
- new effort by OpenAI to solve these problems
- fabricat[ing] authoritative nonsense
- Still true
Our interpretation
- The plain reading is that OpenAI's improved assistant models still sometimes present invented or false factual assertions confidently.
- The 2024 reaffirmation is read as applying to contemporary OpenAI assistants, including GPT-4. The quoted phrase does not uniquely identify a model or intervention.
- The phrase about OpenAI's effort identifies the subject; it is not an antecedent of a conditional forecast.
- For the ongoing observation, use January 1 through March 10, 2024, the latest calendar quarter to date before the post.
Counts as supported if: Documented testing in the target period shows a contemporary OpenAI assistant confidently presenting a demonstrably false factual assertion as an authoritative answer.
Counts as contradicted if: Reliable evaluations establish that the relevant improved OpenAI systems no longer produce such factual fabrications, and alleged examples do not withstand checking.
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
In a January 25, 2024 test, GPT-4 stated that Texas law did not prohibit voters from wearing political apparel at polling places [e1]. The evaluators identified this as false by checking it against Texas's restriction on campaign-related apparel [e1]. This documents a contemporary OpenAI assistant confidently presenting false factual information within the January 1–March 10, 2024 target period, supporting the claim's continuing-occurrence reading [e1].
Limitations
- This example does not establish how frequently such errors occurred across OpenAI products.
- The response does not establish the internal mechanism that produced the false assertion.
- The evidence does not identify the particular OpenAI effort referenced in the older quotation; support follows the contract's reading that includes contemporary assistants such as GPT-4.
Evidence (1 item)
Supports the claim: Proof News and the Institute for Advanced Study, AI Democracy Projects,
Yes, you can wear your MAGA hat to vote in Texas.
Why it matters here: GPT-4 confidently supplied permission and an accompanying false description of Texas polling-place rules, which the evaluators checked against the restriction on campaign apparel. The same report finds GPT-4 substantially more accurate than its peers and describes successful referrals to official polling information.
- Publisher
- Proof News and the Institute for Advanced Study, AI Democracy Projects
- Kind of source
- Independent measurement
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Two years later, deep learning is still faced with the same fundamental challenges
“new effort by OpenAI to solve these problems wound [still] fabricat[ing] authoritative nonsense ” Still true.
- Where
- paragraph 19
- Published
- 2024-03-10
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-two-years-later-deep-learning-is-d4c2a5-a2: 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-two-years-later-deep-learning-is-d4c2a5-a2.