“There is no evidence whatsoever that what has gotten us, sort of, into the game of reading scans is going to bring us all the way into the promised land of a radiologist in an app any time soon.”
Our reading, the sentence that was checked:
There is no evidence that current scan-reading AI techniques will produce a full radiologist replacement any time soon.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Three baffling claims about AI and machine learning in four days, statistical errors in top journals, and claims from Yann LeCun that you should not believe., paragraph 58.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-too-much-benefit-of-the-doubt-c0ff9d-a1
We recorded 13 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2022-10-16. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- “any time soon”
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- There is no evidence whatsoever
- what has gotten us, sort of, into the game of reading scans
- a radiologist in an app
- any time soon
Our interpretation
- Assess the state of publicly available evidence at publication, rather than scoring a forecast against an invented deadline.
- A radiologist in an app ordinarily means a clinically usable replacement for the substantive radiologist role.
- The relevant evidence concerns whether existing scan-reading techniques offered a demonstrated basis for near-term whole-role replacement. Isolated task improvements are relevant but do not automatically establish that conclusion.
- The target is the accumulated public evidence through October 16, 2022, so no arbitrary starting date is imposed.
Counts as supported if: A sufficiently comprehensive contemporary evidence assessment finds no demonstrated basis for near-term whole-role replacement using the scan-reading techniques under discussion.
Counts as contradicted if: Evidence public by the post provides a credible demonstrated path to near-term whole-role replacement using those techniques, rather than only isolated task automation.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
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 search found evidence of autonomous normal-chest-X-ray reporting and research supporting bounded automation. It also found a systematic review describing weak external validation. None establishes imminent whole-role replacement, but these records cannot prove the absolute absence of relevant evidence through October 2022. Whether narrow successes count against 'no evidence whatsoever' depends on an evidential standard and timing Marcus did not supply.
Limitations
- The strongest broad review covers studies only through 2019.
- The autonomous-product announcement is a vendor disclosure.
- Failure to locate whole-role evidence is not proof that none existed.
Evidence (3 items)
Context only: Oxipit,
without any involvement from a radiologist
Why it matters here: The announcement supplies evidence of progress toward autonomous reporting. Its stated scope is normal chest X-rays, so it does not demonstrate replacement of the complete radiologist role.
- Publisher
- Oxipit
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: arXiv; Tommi Keski-Filppula and colleagues,
No critical findings were missed by the software.
Why it matters here: This retrospective study reported successful exclusion of a subset of normal chest X-rays. It is material evidence of bounded automation, but does not evaluate full professional replacement or establish its timing.
- Publisher
- arXiv; Tommi Keski-Filppula and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: European Radiology; Brendan S. Kelly and colleagues,
limited external validation
Why it matters here: The systematic review documents weaknesses in studies published from 2015 through 2019. It supports caution about extrapolation, but cannot establish the absence of relevant evidence through October 2022.
- Publisher
- European Radiology; Brendan S. Kelly and colleagues
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
What would settle it
A dated, comprehensive assessment of evidence available by October 16, 2022, or a contemporary whole-role demonstration with credible clinical validation, evaluated under an explicit interpretation of near-term replacement.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Three baffling claims about AI and machine learning in four days, statistical errors in top journals, and claims from Yann LeCun that you should not believe.
There is no evidence whatsoever that what has gotten us, sort of, into the game of reading scans is going to bring us all the way into the promised land of a radiologist in an app any time soon.
- Where
- paragraph 58
- Published
- 2022-10-16
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-too-much-benefit-of-the-doubt-c0ff9d-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.
About assessment marcus-too-much-benefit-of-the-doubt-c0ff9d-a1.