“Nobody appears to have a technological moat; certainly nobody has a solid answer to the problems that have plagued LLMs for years.”
Our reading:
No LLM company appears to have a technological moat, and none has a solid solution to LLMs' longstanding problems.
Assessment: Can’t be checked
What happened
Not graded on evidence
Outcome: Can’t be checked
An opinion, a recommendation, or too vague to test. Listed, not scored.
Evidence up to . Checked .
Why
The 20 March 2025 preprint reports excessive token generation and reasoning failures in the models tested [e1]. Anthropic's 24 February 2025 disclosure reports reasoning improvements using additional test-time compute [e2]. These contemporaneous findings establish neither an industry-wide absence of solutions nor a durable technological advantage [e1, e2]. The claim cannot be scored as written: 'appears' hedges the moat judgment, and 'solid answer' supplies no task scope or reliability threshold. A verdict would require adding criteria.
Caveats
- The evidence covers selected models and evaluations, not every LLM provider.
- The preprint documents particular limitations; the provider disclosure reports particular improvements. Neither establishes a technological moat or a comprehensive solution.
- The assessment concerns the claim at publication on 27 March 2025, not subsequent progress.
Evidence (2 items)
Claims like this aren’t scored, so there is no evidence to show.
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Where we had to interpret
- The statement concerns the state of LLM providers at publication on 2025-03-27, not their subsequent progress.
- The moat clause is explicitly hedged by 'appears'. It cannot fairly be converted into a definite assertion that no provider possesses a durable technical advantage.
- The context identifies hallucinations and reasoning errors as longstanding problems. However, 'solid answer' does not specify whether a solution means eliminating errors, reducing them substantially, or managing them effectively in particular uses.
- Choosing a duration for a moat or an acceptable error rate would add material criteria the speaker did not supply. The combined claim has no sufficiently determinate test for a binary assessment.
We don’t grade claims like this. We only check the quote and who said it.
We found 21 claims in this post. See them all.
Full record
Source and dates
Marcus on AI, GenAI’s day of reckoning may have come, paragraph 7.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2025-03-27.
Where it was said (1)
Said once in the posts we read.
: Marcus on AI, GenAI’s day of reckoning may have come
Nobody appears to have a technological moat; certainly nobody has a solid answer to the problems that have plagued LLMs for years.
- Where
- paragraph 7
- Published
- 2025-03-27
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Can’t be checked. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
A reviewer said our test wasn’t fair to what was actually said.