“For example, the New York Times reported that Anthropic would have 5GW of capacity by the end of 2026, up from 1.5GW at the end of 2025, according to statements made to its investors, but I can find no compelling evidence that there was even a gigawatt of compute for it to buy.”
Our reading:
Zitron can find no compelling evidence that there was even a gigawatt of compute available for Anthropic to buy, despite New York Times-reported investor statements that it would have 5GW by the end of 2026.
Assessment: Can’t be checked
What happened
Not graded on evidence
Outcome: Can’t be checked
An opinion, a recommendation, or too vague to test. Listed, not scored.
Evidence up to . Checked .
Why
This is a first-person report. Zitron says he 'can find no compelling evidence' that even a gigawatt of compute was available for Anthropic to buy. No one can check what he found or what he would count as compelling. He also does not say when that gigawatt should have existed. His account of the NYT report matches a summary dated 2026-09-18 (e1). That summary says Anthropic told investors it planned to have about five gigawatts by the end of 2026. Anthropic said on 2025-10-23 that its Google TPU expansion was expected to bring 'well over a gigawatt' online in 2026 (e2). On 2026-04-20 Anthropic said it was using over one million Trainium2 chips. It also said nearly 1GW of Amazon capacity would come online by the end of 2026 (e3). Those statements come from the company. The current-use figure is a chip count, not gigawatts. Grading the claim would mean imposing an evidence standard and a date that he did not give.
Caveats
- The NYT original was not read, and the summary (e1) does not confirm the 1.5GW end-of-2025 figure.
- No item gives an independently verified gigawatt figure for Anthropic's operational compute at any date.
- The gigawatt figures in e2 and e3 describe expected capacity, and the current-use statement in e3 is a chip count, not gigawatts.
- The SpaceX Colossus coverage was not read and is not cited.
Evidence (3 items)
Claims like this aren’t scored, so there is no evidence to show.
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Conditions
- I can find no (their words)
Where we had to interpret
- The core claim reports the speaker's own search ('I can find no') against his own standard ('compelling'). Nobody else can check what he found or what he would find compelling.
- The referent of 'there was even a gigawatt of compute for it to buy' is unclear. It could mean new compute available in 2026 to grow from 1.5GW toward 5GW, or compute existing at end of 2025. He gives no date.
- The part about the NYT report can be checked. A summary of the NYT report confirms the 5GW end-of-2026 figure told to investors. The 1.5GW end-2025 figure was not found in the summary.
- Earlier reviewers found that swapping in an objective public-evidence test with a fixed end-2025 date was unfair to the speaker. This contract does not do that.
We don’t grade claims like this. We only check the quote and who said it.
We found 31 claims in this post. See them all.
Full record
Source and dates
Where's Your Ed At, Annualized Enron Rate, paragraph 130.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2026-10-13.
Where it was said (1)
Said once in the posts we read.
: Where's Your Ed At, Annualized Enron Rate
For example, the New York Times reported that Anthropic would have 5GW of capacity by the end of 2026, up from 1.5GW at the end of 2025, according to statements made to its investors, but I can find no compelling evidence that there was even a gigawatt of compute for it to buy.
- Where
- paragraph 130
- Published
- 2026-10-13
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Can’t be checked. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
A reviewer said our test wasn’t fair to what was actually said.