“Large Language Models do not scale profitably, and their functionality is not significant enough to justify the costs of running them.”
Our reading, the sentence that was checked:
LLMs don't scale profitably and their functionality doesn't justify their running costs.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, AI Is A Money Trap, paragraph 128.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-ai-is-a-money-trap-1d3658-a1
We recorded 38 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 2025-08-06. 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
- Large Language Models
- do not scale profitably
- functionality is not significant enough
- justify the costs of running them
Our interpretation
- The ordinary commercial reading is that delivering LLM services at scale cannot earn enough from their useful functionality to cover the costs of providing those services.
- The context makes provider economics central. Customer productivity gains and provider profitability are separate questions.
- Use April through June 2025, the latest completed calendar quarter before the post that public financial reporting could describe.
- The claim names an industry and technology, not a particular company's financial measure; contract metric_scope is therefore null.
- Company losses, serving costs, training costs, and cash burn must be distinguished.
Counts as supported if: Representative service-level evidence for the target period shows that scaled LLM offerings fail to cover properly allocated costs, with evidence that increasing scale does not remedy the shortfall.
Counts as contradicted if: Reliable evidence for the target period shows sustained profitable operation of material LLM services at scale, with the relevant costs included and paying demand attributable to their functionality.
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
The records document financial pressure at individual providers. They do not establish that LLM services generally cannot scale profitably. Anthropic's disclosure also separates customer-serving costs from other expenses, making an inference from company losses to unprofitable inference unsafe. The broad claim remains unsettled for the target quarter.
Limitations
- The primary filing covers one provider.
- OpenAI evidence is a qualified secondary report about a different period and financial measure.
- Later profitability cannot retroactively settle this observation about the period before the post.
Evidence (2 items)
Context only: Anthropic, PBC / U.S. Securities and Exchange Commission,
customer-serving workloads.
Why it matters here: The quarterly table reports a $1.2 billion operating loss for April–June 2025, but one provider's loss cannot establish the industry-wide proposition. Separately, the first-half tables report $1.163 billion of revenue and $1.019 billion of cost of revenue, illustrating why serving economics and overall profitability must be distinguished.
- Publisher
- Anthropic, PBC / U.S. Securities and Exchange Commission
- Kind of source
- Regulatory filing, unaudited
- Measures
- Anthropic, PBC, Operating income (loss), USD millions, gaap, 2025-04-01 to 2025-06-30
- Describes events of
- Public from
Context only: Reuters, republished by Yahoo Finance,
could not immediately verify
Why it matters here: Reuters relayed The Information's report of $2.5 billion in first-half OpenAI cash burn. This is neither a service-level profit measure nor a result confined to the target quarter, and Reuters explicitly qualified its verification.
- Publisher
- Reuters, republished by Yahoo Finance
- Kind of source
- News report
- Measures
- OpenAI, Reported cash burn, USD billions, estimate, 2025-01-01 to 2025-06-30
- Describes events of
- Public from
What would settle it
Comparable target-quarter service accounts showing realized revenue, fully allocated relevant costs, and profitability as usage scales across material LLM offerings.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, AI Is A Money Trap
Large Language Models do not scale profitably, and their functionality is not significant enough to justify the costs of running them.
- Where
- paragraph 128
- Published
- 2025-08-06
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-ai-is-a-money-trap-1d3658-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.