“Some will cost lives, some will cost many lives.”
Our reading, the sentence that was checked:
Some military AI mistakes will cost lives, some many lives.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Is AI already killing people by accident?, paragraph 10.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-is-ai-already-killing-people-by-accident-b70644-a1
We recorded 16 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 2026-03-01. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Forecast
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No deadline stated
- Deadline state on 2026-10-29
- No deadline stated (the release’s reference date)
From the speaker
- Some will cost lives
- some will cost many lives
- AI-triggered mistakes
Our interpretation
- The forecast concerns military AI errors causing deaths after March 1, 2026.
- Both fatal errors and errors causing clearly large losses of life must be established to support the whole claim.
- Many lives has no supplied numerical threshold. A small number of deaths alone cannot establish the larger casualty component.
- AI assistance in a lethal operation is insufficient by itself: an AI error must have contributed to the deaths.
Counts as supported if: Credible investigations establish postpublication military AI mistakes causing deaths, including at least one incident with clearly numerous fatalities.
Counts as contradicted if: Conclusive evidence rules out the predicted events over the forecast's entire scope. A finite period without a confirmed incident cannot contradict this undated forecast.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Did the forecast come true?
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 deadline stated
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
A July autonomous-drone incident reportedly killed three civilians after the aircraft collided with a building. It is relevant evidence, but the accessible account does not conclusively identify an AI error or establish the prediction of errors costing many lives. The Airwars case predates the post and leaves AI involvement uncertain. No qualifying event settles the whole forecast. With no deadline, the absence of stronger evidence by October 28 is not a failure.
Limitations
- Public reporting may omit classified targeting and forensic information.
- The strongest postpublication lead was accessible only through reporting about the original investigation.
- No numerical meaning of many lives was imposed.
Evidence (2 items)
Context only: Tom's Hardware,
The dead were Tetiana Bubynets
Why it matters here: The report describes an autonomously guided drone hitting a building instead of its intended aim point and killing three civilians. This is relevant to the fatal-error component, but does not conclusively isolate an AI error or establish the larger casualty component; no primary forensic report was obtained.
- Publisher
- Tom's Hardware
- Kind of source
- News report
- Describes events of
- Public from
Context only: Airwars,
CENTCOM did not elaborate further.
Why it matters here: Airwars documents a civilian death and inconsistent military statements about whether AI participated in the strike. The incident predates the forecast and does not establish that an AI mistake caused the death.
- Publisher
- Airwars
- Kind of source
- News report
- Describes events of
- Public from
What would settle it
A credible investigation tying a postpublication incident with clearly numerous deaths to a specific military AI error would establish both components.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Is AI already killing people by accident?
Some will cost lives, some will cost many lives.
- Where
- paragraph 10
- Published
- 2026-03-01
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-is-ai-already-killing-people-by-accident-b70644-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-is-ai-already-killing-people-by-accident-b70644-a1.