“Some of this is technical (LLMs aren’t reliable), some of it is political/economic (such as the utter lack of responsible regulation).”
Our reading, the sentence that was checked:
Some of generative AI's downsides have technical causes, such as LLMs being unreliable, and some have political or economic causes, such as the utter lack of responsible regulation.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The growing AI backlash, paragraph 39.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-growing-ai-backlash-4e2b51-a1
We recorded 13 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 2026-05-04. 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
- some of this (scope, stated by the speaker)
From the speaker
- Some of this is technical
- LLMs aren’t reliable
- some of it is political/economic
- the utter lack of responsible regulation
Our interpretation
- Some of this refers to the preceding harms; it is not a conditional forecast antecedent.
- The ordinary causal reading is that model failures contributed to some harms and inadequate governance contributed to some harms.
- The surrounding four-year comparison supplies the cumulative period through publication.
- Responsible regulation concerns adequacy and effectiveness. The existence of a law alone neither establishes adequate protection nor refutes every regulatory-failure claim.
- The categorical phrase utter lack must be considered, but cannot fairly be reduced to a claim that no AI-related law exists anywhere.
Counts as supported if: Documented harms are causally linked both to model unreliability and, in identifiable cases, to missing or ineffective regulatory safeguards.
Counts as contradicted if: Reliable investigations refute the proposed technical or governance mechanisms in the identified cases. Mere existence of legislation is insufficient to refute regulatory inadequacy.
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
The technical component has direct support: the Mata opinion connects fabricated ChatGPT authorities with harm to litigation. It also identifies professional misconduct under existing rules. California had adopted targeted chatbot safeguards before the post, which qualifies the sweeping regulatory language but does not prove regulation was adequate. The causal claim about political or economic regulatory failures remains unsettled under the stated criteria.
Limitations
- One adjudicated incident cannot measure the general reliability of all LLMs.
- The legislative announcement establishes adoption, not effectiveness or the truth of every allegation it describes.
- The schema lacks a government-announcement source type; company_disclosure is used for the official legislative announcement only.
Evidence (2 items)
Supports the claim: US District Court for the Southern District of New York; copy hosted by Relativity,
wastes time and money
Why it matters here: The court found that lawyers submitted nonexistent opinions generated by ChatGPT and described resulting harm. This supports technical unreliability as a contributing cause, while the same opinion assigns responsibility to lawyers under existing rules.
- Publisher
- US District Court for the Southern District of New York; copy hosted by Relativity
- Kind of source
- Regulatory filing
- Describes events of
- Public from
Context only: Office of California State Senator Steve Padilla,
January 1st, 2026.
Why it matters here: This official announcement reports signing SB 243, including chatbot self-harm protocols and a private legal remedy, with operation scheduled before Marcus's post. It qualifies a sweeping description of absent safeguards but does not establish their effectiveness.
- Publisher
- Office of California State Senator Steve Padilla
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
A specified jurisdiction and regulatory standard, plus investigations or comparative evidence linking missing or ineffective safeguards to identified harms.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The growing AI backlash
Some of this is technical (LLMs aren’t reliable), some of it is political/economic (such as the utter lack of responsible regulation).
- Where
- paragraph 39
- Published
- 2026-05-04
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-growing-ai-backlash-4e2b51-a1: 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.