“Coke and Pepsi are commodities; genAI models are, too.”
Our reading, the sentence that was checked:
Generative AI models are commodities.
Assessment: Mixed Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The core misconception that is driving American AI policy, paragraph 12.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-a-deeply-implausible-premise-is-behind-54906a-a2
We recorded 12 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 2025-12-13. 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
- genAI models are, too
- Coke and Pepsi are commodities
- using largely similar products
Our interpretation
- Plain reading: generative AI models from rival providers are broadly interchangeable, so buyers choose mainly on price, and prices are pushed toward cost.
- Present-tense claim: it describes the market in 2025 up to the post date.
Counts as supported if: Evidence that models are easily swapped, that buyers choose mainly on price, and that prices for a given capability fall quickly under competition.
Counts as contradicted if: Evidence that buyers choose mainly on performance, pay for frontier models over cheaper equivalents, and stick with particular providers, so models are differentiated rather than interchangeable.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Mixed
Part of the claim holds and part does not, and the record says which part is which.
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 contract reads 'commodity' as meaning that rival models are easy to swap, that buyers choose mainly on price, and that prices for a given capability fall quickly under competition. Evidence of buyers choosing on performance, paying for frontier models and staying with one provider would contradict it. Part of the claim holds. Epoch AI (March 12, 2025) found that the price of reaching a given benchmark result fell 9x to 900x per year (e1). Menlo Ventures' mid-2025 report says switching vendors is relatively easy (e3). Part does not hold. The same report says builders 'consistently choose frontier models over cheaper, faster alternatives' and 'pay for performance' (e2). It also reports that 66% of builders upgraded within their existing provider, 23% did not switch models, and only 11% switched vendors (e3). These buyer findings match the contract's contradiction condition. The price findings match part of the support condition. The evidence does not show whether prices were pushed toward cost.
Limitations
- The buyer evidence comes from a single survey of builders by a venture firm, from mid-2025, months before the December 2025 post.
- Epoch measured price declines at benchmark milestones. It did not measure whether buyers treat models as interchangeable, and it found no clear evidence that falling margins explain the declines.
- There are no profit-margin data showing whether prices reached cost.
- 'Commodity' is partly a rhetorical label, so the result depends on the reading used.
Evidence (3 items)
Supports the claim: Epoch AI,
The rate of decline varies dramatically depending on the performance milestone, ranging from 9x to 900x per year.
Why it matters here: The price of any given capability level falls very fast. That is consistent with commodity-like competition once a capability is matched.
- Publisher
- Epoch AI
- Kind of source
- Independent measurement
- Describes events of
- Public from
Goes against the claim: Menlo Ventures,
Performance drives decisions. Builders consistently choose frontier models over cheaper, faster alternatives.
Why it matters here: Buyers paid for top performance rather than treating models as interchangeable and choosing on price. That cuts against the commodity claim for frontier models.
- Publisher
- Menlo Ventures
- Kind of source
- Independent measurement
- Describes events of
- Public from
Goes against the claim: Menlo Ventures,
66% of builders upgraded models within their existing provider, while 23% did not switch models at all this past year. Only 11% switched vendors.
Why it matters here: Few buyers switched vendors. That points to provider loyalty, not commodity-style swapping, although the same report says switching is 'relatively easy.'
- Publisher
- Menlo Ventures
- Kind of source
- Independent measurement
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The core misconception that is driving American AI policy
Coke and Pepsi are commodities; genAI models are, too.
- Where
- paragraph 12
- Published
- 2025-12-13
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-a-deeply-implausible-premise-is-behind-54906a-a2: Current. Outcome: Mixed. 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.
About assessment marcus-a-deeply-implausible-premise-is-behind-54906a-a2.