“First, the language technology that the new system relies on is well-known to be problematic and second it is likely to be even more problematic in the context of robots.”
Our reading, the sentence that was checked:
The language technology that PaLM-SayCan relies on is well known to be problematic.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, What’s wrong with Google’s new robot project, paragraph 4.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-whats-wrong-with-googles-new-robot-730ea2-a1
We recorded 20 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 2022-08-18. 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
- the language technology that the new system relies on
- well-known to be problematic
Our interpretation
- Assess the normalized observation about language models, excluding the separate prediction that embodiment would make their problems worse.
- The plain reading is that significant limitations and risks were already documented and recognized in research, not that language models were useless.
- Use the latest research reporting year to date, 2022-01-01 through 2022-08-18. Research has no standard quarterly reporting period; earlier publications provide background.
Counts as supported if: Contemporary primary research documents significant language-model problems, with acknowledgment by the developers of PaLM or other evidence that these problems were recognized beyond an isolated example.
Counts as contradicted if: Contemporary research establishes that the alleged problems were absent or already reliably resolved in the relevant language technology.
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
Independent research documented toxic generation, and PaLM's developers acknowledged the existing literature on language-model risks. That supports the ordinary reading that significant problems were already recognized by August 2022. It does not establish that every language-model application was unreliable or that embodiment necessarily worsened the risks.
Limitations
- Recognition is established within the research community, not through a survey of public awareness.
- This assessment excludes the separate prediction about greater problems in robots.
Evidence (3 items)
Supports the claim: Google Research,
various potential risks
Why it matters here: PaLM's own announcement acknowledges an existing research literature on risks and describes analyses of bias and model outputs. This directly supports recognition of problems in the relevant technology.
- Publisher
- Google Research
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Supports the claim: Proceedings of the AAAI Conference on Artificial Intelligence,
toxic language
Why it matters here: The published research documents a concrete problem rather than merely forecasting possible harm. Its proposed mitigation does not establish that toxic generation had been eliminated.
- Publisher
- Proceedings of the AAAI Conference on Artificial Intelligence
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: Google DeepMind,
risk mitigations
Why it matters here: DeepMind's earlier discussion identifies misinformation, stereotypes, and other risks while acknowledging unfinished mitigation work. It supplies historical background outside the selected reporting period.
- Publisher
- Google DeepMind
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, What’s wrong with Google’s new robot project
First, the language technology that the new system relies on is well-known to be problematic and second it is likely to be even more problematic in the context of robots.
- Where
- paragraph 4
- Published
- 2022-08-18
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-whats-wrong-with-googles-new-robot-730ea2-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-whats-wrong-with-googles-new-robot-730ea2-a1.