He thinks it is noteworthy that OpenAI and Anthropic either wildly miss their cost projections or almost exactly hit their revenue projections, and that the latter is even weirder.

Not graded on evidence

Outcome: Can’t be checked

An opinion, a recommendation, or too vague to test. Listed, not scored.

Evidence up to 2026-10-28. Checked 2026-10-31.

Why

The sentence mostly gives the speaker's opinion. He calls something a 'noteworthy story' and 'even weirder.' Its factual part is a loose pattern: cost projections are wildly missed, while revenue projections are almost exactly met. He does not name a projection, a date or a year. He gives no threshold for 'wildly' or 'almost exactly.' He does not say whether a miss means actual results against a projection, or later projections against earlier ones. He also does not say whether the pattern has to hold for both OpenAI and Anthropic. Without these details, any test would rest on comparisons the evaluator picks for him. So the claim cannot be fairly graded.

Caveats

• No source was read for actual revenue or cost figures. The packet holds only search summaries, and they are not outcome evidence.
• This assessment says nothing about whether any particular OpenAI or Anthropic projection was met or missed.
• A narrower claim naming specific projections, periods and thresholds could be checked.

Evidence (0 items)

Claims like this aren’t scored, so there is no evidence to show.