“What ChatGPT delivered, in spades, that its predecessors like Microsoft Tay (released March 23, 2016, withdrawn March 24 for toxic behavior) and Meta’s Galactica (released November 16, 2022, withdrawn November 18) could not, was an illusion—a sense that the problem of toxic spew was finally coming under control.”
Our reading, the sentence that was checked:
ChatGPT's apparent control of toxic output is an illusion rather than a real solution.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Inside the Heart of ChatGPT’s Darkness, paragraph 2.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-inside-the-heart-of-chatgpts-darkness-27ccf8-a1
We recorded 11 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 2023-02-11. 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
- ChatGPT
- an illusion
- a sense that the problem of toxic spew was finally coming under control
Our interpretation
- The ordinary reading is that apparent safety did not withstand adversarial prompting. It is not that safeguards produced no improvement: Marcus expressly acknowledges frequent refusals.
- Assess the current partial month, February 1–11, 2023, because the surrounding discussion concerns current behavior and examples from that week.
- The Tay and Galactica parentheticals are background, not the proposition selected for assessment.
Counts as supported if: Contemporaneous independent reports document successful prompt-based circumvention of ChatGPT's restrictions on toxic or harmful output.
Counts as contradicted if: Contemporaneous reproducible tests show that the alleged circumvention did not produce the claimed outputs, and substantial adversarial testing supports the apparent robustness.
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
Contemporary reporting documents prompt-based access to restricted content. This supports the contextual reading that apparent safety was vulnerable to circumvention. The same report acknowledges failures, so it does not establish that safeguards achieved nothing. The result depends on reading 'illusion' as a claim about robustness.
Limitations
- The decisive source is contemporary secondary reporting, not a controlled benchmark.
- No finding about the complete absence of safety improvements or public perceptions is proposed.
Evidence (1 item)
Supports the claim: Dark Reading,
allowing DAN to bypass content restrictions in ChatGPT.
Why it matters here: The report describes a currently available method for bypassing restrictions on violent and illegal content. It also says the method does not always succeed, limiting any inference that safeguards were wholly ineffective.
- Publisher
- Dark Reading
- Kind of source
- News report
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Inside the Heart of ChatGPT’s Darkness
What ChatGPT delivered, in spades, that its predecessors like Microsoft Tay (released March 23, 2016, withdrawn March 24 for toxic behavior) and Meta’s Galactica (released November 16, 2022, withdrawn November 18) could not, was an illusion—a sense that the problem of toxic spew was finally coming under control.
- Where
- paragraph 2
- Published
- 2023-02-11
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-inside-the-heart-of-chatgpts-darkness-27ccf8-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-inside-the-heart-of-chatgpts-darkness-27ccf8-a1.