“And before then, you’re going to get a lot of desperate stories about how “AI gains can be found outside of training new models” to try and keep the party going, despite reality flicking the lights on and off and threatening to call the police.”
Our reading, the sentence that was checked:
Before hyperscalers pull back, there will be many desperate stories claiming AI gains can come from outside training new models.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Godot Isn't Making it, paragraph 87.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-godot-isnt-making-it-640fc9-a2
We recorded 33 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 2024-12-03. 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)
Conditions
- before hyperscalers pull back (qualifier, stated by the speaker)
From the speaker
- before then
- a lot of desperate stories
- AI gains can be found outside of training new models
- to try and keep the party going
Our interpretation
- 'Before then' refers to the preceding prediction of a Google, Microsoft, or Meta capital-spending pullback.
- The observable content is repeated, independently published coverage after the post arguing that AI performance can improve through inference-time computation or other methods beyond training new models.
- 'A lot' means a recurring theme across multiple independent outlets, not one isolated story or repeated syndication of the same article. The speaker supplied no numerical threshold.
- The supplied condition is assessed as an ordering requirement: coverage must precede a verified qualifying pullback. Finding coverage alone does not establish the complete sequence.
- 'Desperate' and 'keep the party going' characterize tone and motive. Reporting on inference-time methods alone does not establish either.
Counts as supported if: A dated body of independent post-publication coverage establishes the recurring theme, and a verified later pullback establishes the required ordering. A finding about desperation or promotional intent requires separate evidence beyond the topic of coverage.
Counts as contradicted if: A verified qualifying pullback closes the relevant interval, and sufficiently comprehensive coverage records show that the predicted recurring theme did not appear before it. Isolated skeptical stories do not by themselves contradict the 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
Scientific American discussed test-time computation as an alternative route to model improvement on January 23, 2025 [e5]. TechCrunch's March 19, 2025 article, updated March 20, covered inference-time search using an existing model and included expert skepticism [e6]. These two independent articles establish that the predicted topic appeared, but do not establish abundant coverage or the claimed tone and motive [e5, e6]. Neither article establishes a qualifying later capital-spending pullback, so they do not settle the required sequence [e5, e6]. The forecast remains unresolved.
Limitations
- Two verified examples are not a systematic record establishing the quantity described as 'a lot'.
- The packet does not verify a qualifying Google, Microsoft or Meta capital-spending pullback or its date.
- Coverage of inference-time methods does not itself establish desperation or promotional intent.
Evidence (2 items)
Supports the claim: Scientific American,
test-time compute
Why it matters here: This directly supports the coverage component: a post-publication story presented test-time computation as an alternative route to improvement. Its discussion of costs and limits does not establish desperation or an intent to sustain hype.
- Publisher
- Scientific American
- Kind of source
- News report
- Describes events of
- Public from
Supports the claim: TechCrunch,
test-time scaling
Why it matters here: A second independent outlet covered inference-time search and claimed gains using an existing model. The article also presented expert criticism, which cautions against treating such coverage as evidence of a desperate promotional motive.
- Publisher
- TechCrunch
- Kind of source
- News report
- Describes events of
- Public from
What would settle it
Verify a qualifying capital-spending pullback and its date, then assess a sufficiently broad record of independent articles published after December 3, 2024 and before that event. Assess the claimed tone and motive separately using evidence beyond the articles' technical subject matter.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Godot Isn't Making it
And before then, you’re going to get a lot of desperate stories about how “AI gains can be found outside of training new models” to try and keep the party going, despite reality flicking the lights on and off and threatening to call the police.
- Where
- paragraph 87
- Published
- 2024-12-03
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-godot-isnt-making-it-640fc9-a2: 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.