“To be clear, OpenAI burned approximately $5 billion on compute alone.”
Our reading, the sentence that was checked:
OpenAI spent about $5 billion on compute alone in 2024.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, The Generative AI Con, paragraph 32.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-longcon-501d82-a2
We recorded 30 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 2025-02-17. 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
- OpenAI
- approximately $5 billion
- on compute alone
Our interpretation
- The surrounding discussion of last year's losses supplies calendar 2024.
- Compute means company-wide training, research, and inference compute costs, including noncash compute costs.
- For a numerical test, approximately $5 billion means an amount that rounds to $5 billion to the nearest billion. This tolerance is my interpretation, not a threshold supplied by Zitron.
- The passage relies on press estimates and does not specify cash expenditure versus expense recognition.
Counts as supported if: A reconciled full-year estimate places total 2024 compute costs between $4.5 billion and $5.5 billion on the stated inclusive basis.
Counts as contradicted if: Reliable full-year evidence places comparable total compute costs materially outside that interval, with enough precision to overcome reporting and accounting uncertainty.
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
Epoch AI's October 10, 2025 analysis estimates OpenAI's 2024 cloud compute costs at about $7 billion, above the claimed approximately $5 billion [e1]. The estimate partly relies on fourth-quarter projections and an assumed research-cost amortization schedule [e1]. It raises doubt but does not establish a comparable actual full-year total precise enough to confirm or contradict the claim [e1].
Limitations
- The packet contains no verified full-year reconciliation of training, research and inference compute costs.
- Costs incurred, recognized expenses and cash payments are not interchangeable.
- The $4.5–$5.5 billion comparison interval is an evaluator interpretation of 'approximately.'
Evidence (1 item)
Context only: Epoch AI,
OpenAI spent ~$7B on cloud compute last year—mostly R&D (research, experiments, training).
Why it matters here: This estimate is higher than the claim, but it includes an assumed research-cost amortization schedule. Epoch explicitly says the underlying investor figures rely partly on projections for the final quarter of 2024, so this is not decisive evidence of actual comparable spending.
- Publisher
- Epoch AI
- Kind of source
- Secondary summary
- Measures
- OpenAI, estimated total cloud compute costs incurred, with research amortization reversed, USD billions, estimate, 2024-01-01 to 2024-12-31
- Describes events of
- Public from
What would settle it
A reconciled actual 2024 total covering training, research and inference compute without double counting, with cash payments, noncash costs, recognized expenses and amortization distinguished.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, The Generative AI Con
To be clear, OpenAI burned approximately $5 billion on compute alone.
- Where
- paragraph 32
- Published
- 2025-02-17
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-longcon-501d82-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.