“Your private data are going to feed the insatiable beast, too — like it or not.”
Our reading, the sentence that was checked:
People's private data will be used to feed data-hungry AI systems, whether or not they like it.
Assessment: Supported Settled before its deadline on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The dirty underbelly of AI, paragraph 10.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-dirty-underbelly-of-ai-5d0270-a2
We recorded 2 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 2023-09-07. 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
- Settled before its deadline
- Deadline state on 2026-10-29
- Settled before its deadline (the release’s reference date)
Conditions
- whether or not people consent (qualifier, stated by the speaker)
From the speaker
- Your private data
- are going to feed
- like it or not
- The paradigm we have right now
Our interpretation
- The ordinary reading is that AI developers will use people's nonpublic information for training without requiring affirmative permission from each affected person.
- Private data includes nonpublic chatbot conversations and uploaded personal material. Public social-media posts alone do not establish this claim.
- Default training with an opt-out qualifies as use without affirmative opt-in. It does not establish that a company ignored an exercised opt-out.
- The statement predicts a general practice, rather than claiming that every person's private information will inevitably be used.
- The phrase 'like it or not' describes how data will be used; it is not a separate conditional antecedent.
Counts as supported if: A dated disclosure or reliable investigation after publication establishes that a major AI developer uses nonpublic user information for model training without requiring affirmative opt-in permission.
Counts as contradicted if: Decisive evidence establishes that the predicted practice cannot occur under the described AI paradigm. An individual company's opt-out or exclusion policy limits the claim's breadth but cannot disprove this open-ended forecast.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Did the forecast come true?
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
- Settled before its deadline
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
Google's October 26, 2026 FAQ says reviewers assess a subset of nonpublic Gemini chats and that their suggestions help improve its models [e1]. A policy study published October 15, 2025 identifies Google as using chat inputs for training by default, with an opt-out [e2]. Together, these records support the forecast under the contract's reading of use without affirmative opt-in [e1, e2]. Google's August 13, 2025 disclosure provides training exclusions through settings and Temporary Chats, so the evidence does not establish unavoidable use for everyone or disregard of an exercised opt-out [e3].
Limitations
- The assessment treats default participation as sufficient; it does not establish that Google ignored an exercised opt-out [e1, e2, e3].
- The policy study examines disclosed practices rather than underlying training datasets [e2].
- Google's disclosed exclusions limit any claim that everyone's private information must be used [e3].
Evidence (3 items)
Supports the claim: Google,
help our models respond better
Why it matters here: Google describes reviewers assessing users' nonpublic conversations and supplying improvements used by its models. The same FAQ explicitly includes generative-model training among uses of saved activity when Keep Activity is enabled; e2 establishes the default-participation policy.
- Publisher
- 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/ACM Conference on AI, Ethics, and Society,
training by default
Why it matters here: Table 1 records both default training on chat inputs and an opt-out mechanism for Google. This independent policy analysis corroborates that training participation does not require an affirmative opt-in, although it does not inspect individual training datasets.
- Publisher
- Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: Google,
turn this setting off
Why it matters here: Google provides an opt-out and temporary conversations excluded from training. This limits any interpretation that every user's information must be used or that an exercised objection is necessarily ignored.
- Publisher
- 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, The dirty underbelly of AI
Your private data are going to feed the insatiable beast, too — like it or not.
- Where
- paragraph 10
- Published
- 2023-09-07
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-dirty-underbelly-of-ai-5d0270-a2: 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-the-dirty-underbelly-of-ai-5d0270-a2.