“Within hours of when I posted yesterday’s essay on the decline of tokenmaxxing, I saw further converging evidence, such as this:”
Our reading, the sentence that was checked:
There is further converging evidence, as of late May 2026, that tokenmaxxing (the push to maximize AI token usage) is declining.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, What happens next, after the decline of tokenmaxxing?, paragraph 1.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-what-happens-next-after-the-decline-778ba6-a2
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2026-05-29. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- Within hours
- yesterday’s essay on the decline of tokenmaxxing
- further converging evidence
- such as this
Our interpretation
- The claim concerns publicly observable evidence of retreat from encouraging token consumption for its own sake, rather than Marcus's personal experience of seeing reports.
- The plain reading is that several developments corroborated an emerging retreat. It does not require tokenmaxxing to have ended or aggregate AI usage to have fallen.
- I use May 21–29, 2026 for the ongoing late-May state described. Earlier developments are context unless the evidence establishes a relevant state during this period.
- Removing incentives for unnecessary usage directly bears on the claim. Cost concerns, changing suppliers, and GPU rental prices require an additional connection to reduced encouragement of tokenmaxxing.
Counts as supported if: Multiple independently documented developments during the target period show organizations withdrawing incentives for maximizing token consumption or explicitly restraining unnecessary usage.
Counts as contradicted if: Reliable contemporaneous evidence disproves the purported retreats or establishes that the cited developments maintained or intensified incentives to maximize token consumption. Continued growth in useful AI adoption alone would not contradict the claim.
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
On May 29, 2026, Amazon confirmed that KiroRank had been deprecated; the report also quoted management guidance that week against using AI for its own sake (e1). This establishes one qualifying retreat, but the May 28 Fortune report does not date Meta's leaderboard removal (e1, e2). Microsoft's change was reported on May 14, before the May 21–29 target period, and retained access to Anthropic models through Copilot CLI (e3). These records do not establish multiple independent retreats during the target period, but they do not disprove such convergence either (e1–e3).
Limitations
- Amazon's precise leaderboard shutdown date is unspecified (e1).
- Meta's removal date and any change in Microsoft's incentives for total token consumption are not established (e2, e3).
- The packet lacks a verified comparable H200 rental-price series or an established connection between those prices and tokenmaxxing incentives.
Evidence (3 items)
Supports the claim: Business Insider, republished by Yahoo Finance,
had been "deprecated."
Why it matters here: Amazon confirmed that the leaderboard was no longer operating by this date; the precise shutdown day is unspecified. This establishes one concrete retreat from a mechanism encouraging excessive usage, although Amazon described it as informal and employee-created.
- Publisher
- Business Insider, republished by Yahoo Finance
- Kind of source
- News report
- Describes events of
- Public from
Context only: Fortune,
cancelled Claude Code subscriptions
Why it matters here: Reported restriction.
- Publisher
- Fortune
- Kind of source
- News report
- Describes events of
- Public from
Context only: The Verge,
remain accessible through Copilot CLI
Why it matters here: The original report describes replacement access, limiting what the subscription change establishes about tokenmaxxing. It also predates the target window and therefore serves as context.
- Publisher
- The Verge
- Kind of source
- News report
- Describes events of
- Public from
What would settle it
Dated, independent records establishing another withdrawal of usage-maximizing incentives or explicit restraint of unnecessary usage during May 21–29, 2026; alternatively, reliable contemporaneous evidence disproving the purported retreats or showing that they preserved or intensified those incentives.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, What happens next, after the decline of tokenmaxxing?
Within hours of when I posted yesterday’s essay on the decline of tokenmaxxing, I saw further converging evidence, such as this:
- Where
- paragraph 1
- Published
- 2026-05-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-what-happens-next-after-the-decline-778ba6-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.
About assessment marcus-what-happens-next-after-the-decline-778ba6-a2.