“The problem with this argument is that we’re yet to prove if running these models is profitable (or even sustainable) for any provider, nor do we have tangible proof that they can compete at scale with Anthropic or OpenAI’s more-complex LLMs.”
Our reading, the sentence that was checked:
It is not yet proven whether running cheaper open-source models is profitable, or even sustainable, for any provider, and there is no tangible proof that they can compete at scale with Anthropic's or OpenAI's more-complex LLMs.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, AI's Brokenomics, paragraph 108.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-brokenomics-9c481e-a1
We recorded 41 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 2026-06-15. 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
- we’re yet to prove
- running these models is profitable (or even sustainable)
- for any provider
- nor do we have tangible proof
- compete at scale
- Anthropic or OpenAI’s more-complex LLMs
Our interpretation
- This concerns the public evidence available by June 15, 2026, rather than whether profitability or competitive deployment appeared later.
- These models refers to the cheaper open-source or open-weight alternatives discussed immediately before the statement.
- Profitable running means revenue from hosting the models covers the costs of providing that service. Revenue growth, financing, and hypothetical billing do not establish this.
- Competing at scale means credible production evidence of useful substitution at meaningful volume, with relevant quality, reliability, and cost comparisons.
- The context concerns routing some tasks to cheaper models, so the test does not demand superiority on every frontier task.
- No accounting basis, scale threshold, or duration for sustainability is supplied.
Counts as supported if: A sufficiently broad examination of the public record establishes that neither actual profitable hosting nor meaningful competitive production operation had been demonstrated by publication. Failure to find evidence in a limited search alone is insufficient.
Counts as contradicted if: A prepublication public record demonstrates actual profitable hosting by a provider, or a credible production comparison demonstrates competitive substitution at meaningful scale. Each limb must be assessed separately.
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
There was tangible evidence of open-model production use and a published comparison with a GPT-4 product feature. Those records challenge a sweeping reading that no practical competitive evidence existed. They do not jointly establish a matched frontier-model comparison at scale. DeepSeek's advertised economics were hypothetical and do not establish realized profit. The search also cannot prove that no qualifying provider evidence existed anywhere. Neither the full claim nor its negation is established.
Limitations
- The result depends partly on the undefined meaning of compete at scale.
- Fireworks supplies a vendor disclosure, not independently verified comparative measurements.
- The older product study does not test the frontier alternatives available at publication.
- Unknown accounting basis prevents a strong financial score.
- Evidence published after June 15 cannot refute an assertion about what had been demonstrated by that date.
Evidence (3 items)
Context only: DeepSeek,
$562,027
Why it matters here: The calculation assumes all usage is billed at R1 prices; the document says actual revenue is lower because of cheaper models, free access, and discounts. It therefore cannot establish realized hosting profit.
- Publisher
- DeepSeek
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Measures
- DeepSeek, Hypothetical daily inference revenue if all usage were billed at R1 prices, USD dollars per day, estimate, 2025-02-27 to 2025-02-28
- Describes events of
- Public from
Context only: Fireworks AI,
millions of users worldwide
Why it matters here: The provider reports substantial production adoption of its open-model platform. This is material evidence against a sweeping absence-of-deployment reading, but it does not provide a matched frontier-model comparison or hosting profit accounts.
- Publisher
- Fireworks AI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: IEEE ISPASS 2024 / arXiv,
GPT-4 API
Why it matters here: This is concrete comparative evidence for a particular product feature, rather than merely a generic benchmark. Its older comparator and limited deployment scope do not establish the broader mid-2026 scale comparison.
- Where
- Abstract
- Publisher
- IEEE ISPASS 2024 / arXiv
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
What would settle it
A dated prepublication hosting-profit disclosure on an identified basis, or a matched production comparison showing useful substitution at meaningful scale. A broader documented review would be needed to support the asserted absence of both.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, AI's Brokenomics
The problem with this argument is that we’re yet to prove if running these models is profitable (or even sustainable) for any provider, nor do we have tangible proof that they can compete at scale with Anthropic or OpenAI’s more-complex LLMs.
- Where
- paragraph 108
- Published
- 2026-06-15
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-brokenomics-9c481e-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.