“It has no moat, its models are hitting the point of diminishing returns and have been for some time, and as popular as ChatGPT may be, it isn't a business and constantly loses money.”
Our reading, the sentence that was checked:
OpenAI's models are hitting diminishing returns and have been for some time.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Desperate Times, Desperate Measures, paragraph 64.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-measures-fe076f-2-a1
We recorded 15 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 2025-05-27. 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
- its models
- hitting the point of diminishing returns
- have been for some time
Our interpretation
- The ordinary reading is that successive OpenAI models were delivering smaller improvements relative to the resources required to develop and run them.
- The claim covers OpenAI's model development through the post date, including reasoning models. It does not expressly restrict itself to pretraining.
- The starting date is unspecified. I leave it open rather than inventing a duration for 'some time'.
- Improving benchmark scores alone neither establish nor disprove diminishing returns.
Counts as supported if: Comparable measurements across successive OpenAI models show a sustained decline in useful performance gained per additional training or inference resource.
Counts as contradicted if: Comparable measurements across successive OpenAI models show sustained or increasing marginal performance gains for comparable additional resources over the period described.
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
GPT-4.5's resource burden supports the concern behind the claim. Independent o3 testing and OpenAI's reasoning-model comparisons show substantial gains and improved inference economics. These observations can coexist with diminishing marginal returns. The available evidence does not establish or refute a sustained resource-adjusted trend across OpenAI's models.
Limitations
- Company-reported cost comparisons are not independent measurements of total development resources.
- A single benchmark or model transition cannot establish the claimed sustained trend.
Evidence (3 items)
Context only: OpenAI,
GPT‑4.5 is a very large and compute-intensive model, making it more expensive than and not a replacement for GPT‑4o.
Why it matters here: This is consistent with concerns about expensive improvements. It supplies neither a quantified performance gain nor a series establishing diminishing returns.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: ARC Prize,
OpenAI's new o3 model represents a significant leap forward in AI's ability to adapt to novel tasks.
Why it matters here: Independent testing challenges a reading of general capability stagnation. The tested system was an o3 preview, and a capability leap does not by itself establish increasing returns on total resources.
- Publisher
- ARC Prize
- Kind of source
- Independent measurement
- Describes events of
- Public from
Context only: OpenAI,
For example, on the 2025 AIME math competition, the cost-performance frontier for o3 strictly improves over o1, and similarly, o4-mini's frontier strictly improves over o3‑mini.
Why it matters here: This is material counterevidence to a blanket claim of worsening inference economics. One company-reported task comparison does not settle the broader marginal-return trend, including training resources.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
A consistent historical comparison of successive OpenAI models, measuring useful performance gains against training and inference resources through May 27, 2025.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Desperate Times, Desperate Measures
It has no moat, its models are hitting the point of diminishing returns and have been for some time, and as popular as ChatGPT may be, it isn't a business and constantly loses money.
- Where
- paragraph 64
- Published
- 2025-05-27
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-measures-fe076f-2-a1: 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.