“But this is nonsense; even a few seconds with Google Images supplies lots of people drawing with their left hand.”
Our reading:
The critic's objection that the system had not been taught relevant information about left-handed drawing is nonsense, because a few seconds on Google Images turns up lots of people drawing with their left hand.
Assessment: Unresolved
What happened
Was it true for the period described?
Outcome: Unresolved
We checked, and the evidence so far doesn’t settle it.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
The claim requires more than showing that left-handed drawing images exist. Marcus says a quick Google Images search supplies many examples and uses this to dismiss an objection about the system's training. The searches did not locate independent records of the relevant Google results or evidence establishing the system's exposure to the relevant information. Public discoverability would not by itself establish that exposure. Neither component is settled by the retrieved, citable evidence, so the whole claim remains unresolved.
Caveats
- Historical search results were not recovered; present-day results cannot establish the claimed experience in February 2024.
- Several potentially relevant sources could not be read or verified with read_page.
- No adverse inference is drawn from unavailable training records or unsuccessful retrievals.
Evidence (0 items)
Nothing we found settles this yet.
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Conditions
- rebuttal to one critic's objection about the system (their words)
Where we had to interpret
- The claim combines a factual premise about readily discoverable images with a rebuttal to the objection about the system's training. Support requires evidence for both.
- I read 'a few seconds' as a quick, ordinary search, and 'lots' as numerous distinct examples readily visible without extended searching. Marcus supplies no exact time or count.
- The relevant images must depict people drawing with their own left hand, rather than merely depict drawings of hands.
- The rebuttal requires evidence that the system had received relevant information about left-handed drawing. It does not require proof that every Google result was included in training, or that the system reliably understood handedness.
- The immediate context describes a search possible at publication. I therefore use February 13, 2024 as the target date.
- The description of a critic's objection supplies argumentative context, not a conditional forecast.
Right if: Contemporaneous search records or a credible historical reconstruction show numerous distinct images of people drawing with their left hand readily discoverable through an ordinary Google Images search on the target date. Evidence about the system's actual training also establishes that it received relevant information about left-handed drawing, defeating the objection as Marcus describes it. Public availability alone does not satisfy this second requirement.
Wrong if: Reliable historical search evidence shows that ordinary relevant Google Images searches did not readily return numerous such examples, or reliable evidence about the identified system establishes that it had not received the relevant information. Missing search archives or undisclosed training data alone do not establish contradiction.
We wrote this test after the outcome was known.
What would settle it
Dated Google Images captures or a credible historical reconstruction showing the relevant results, together with identification of the system and reliable evidence about its exposure to left-handed drawing information. Evidence establishing that either factual component was false could instead contradict the claim.
We found 7 claims in this post. See them all.
Full record
Source and dates
Marcus on AI, Statistics versus Understanding: The Essence of What Ails Generative AI, paragraph 14.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-29. The post may have been edited after 2024-02-13.
Where it was said (1)
Said once in the posts we read.
: Marcus on AI, Statistics versus Understanding: The Essence of What Ails Generative AI
But this is nonsense; even a few seconds with Google Images supplies lots of people drawing with their left hand.
- Where
- paragraph 14
- Published
- 2024-02-13
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Unresolved. 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.