Large language models exhibit the traits of a general-purpose technology
Assessment
Evidence favors the claim, but the chain is incomplete or the sources are secondary.
Economists identify a general-purpose technology by three traits: pervasive applicability across sectors, sustained improvement over time, and the spawning of complementary innovations. Large language models show each trait to a degree unusual for so young a technology. They are applicable to tasks across a wide range of occupations and industries, from software and law to medicine, education, and research; their capabilities have improved rapidly on benchmarks, cost, and breadth of use since 2020; and a substantial ecosystem of complementary products and workflows has been built on top of them.
The characterization originates with Eloundou and colleagues' 2023 study and has since been echoed, in hedged form, by an OECD review of the early evidence, which finds generative AI has considerable potential to qualify as a general-purpose technology. The main caution is timing rather than substance: general-purpose technology status is normally confirmed in retrospect, after decades of diffusion and productivity effects, and skeptics such as Daron Acemoglu argue that the share of work where these models are genuinely cost-effective, and hence their aggregate economic impact, will prove much smaller than early exposure estimates suggest. That dispute concerns the magnitude of eventual impact more than the presence of the traits themselves, which is what this claim asserts. On the evidence to date the traits are present; whether they mature into transformation on the scale of electricity or computing remains open.
Full reasoning: the evidence and decisions behind this verdict
The claim comes from Eloundou, Manning, Mishkin, and Rock, "GPTs are GPTs" (arxiv.org/abs/2303.10130), whose abstract asserts it directly. The verdict rests on the three-criteria test from the general-purpose technology literature (Bresnahan and Trajtenberg), examined trait by trait.
Pervasiveness: occupational exposure studies (Eloundou et al.; Felten et al.) find LLM-relevant tasks across most of the occupation distribution, and deployment is observable across many sectors. The applicability subclaim is seeded high; the honest caveat is that exposure measures potential applicability, and Acemoglu's work (e.g., his projection that generative AI will automate roughly 4.6% of tasks within a decade, shass.mit.edu/external-links/daron-acemoglu-thinks-ai-is-solving-the-wrong-problems) argues cost-effective applicability is far narrower.
Improvement: rapid capability improvement is well documented in benchmark trajectories, falling inference costs, longer contexts, and multimodality; the live debate about diminishing returns to scaling bears on the future rate, not the record to date.
Complementary innovation: the innovation-spawning subclaim is the most directly observable: coding assistants, agent frameworks, retrieval systems, and vertical applications built on LLM APIs constitute exactly the co-invention the criterion describes, though its durability is not yet proven.
Instance stances run one way: Eloundou et al. affirm; the OECD paper by Calvino, Haerle, and Liu affirms in hedged form ("considerable potential to qualify", www.oecd.org/en/publications/is-generative-ai-a-general-purpose-technology_704e2d12-en.html); an MIT Sloan report treats the characterization as given. A targeted search found no credible source asserting the negation; skepticism in the discourse (Acemoglu, "great AI disappointment") targets economic magnitude and hype, not the trait test. That neighboring dispute is genuinely live but belongs to claims about productivity impact, where the graph separately holds that general-purpose technologies historically took decades before producing measurable aggregate productivity gains, a point that cuts against inferring "not a GPT" from small measured gains so far.
The verdict is supported rather than verified because the claim is partly a forward-looking economic characterization: GPT status has historically been confirmed retrospectively, the strongest affirmations beyond the originating paper are hedged, and the pervasiveness trait rests substantially on exposure estimates rather than realized adoption. What would move the verdict: several more years of adoption and productivity data (toward verified if diffusion deepens across sectors; toward contested if adoption stalls in a narrow set of use cases), or a credible economic analysis directly arguing the traits are absent.
Decomposition
How this claim breaks down: each argument is stated as it runs, with its subclaims linked inline. ↗︎ opens a subclaim; the map shows how they fit together.
Economists identify a general-purpose technology by three traits: pervasive applicability across sectors, continual improvement over time, and the spawning of complementary innovations. Because large language models are applicable across a wide range of occupations and industries, their capabilities are improving rapidly over time, and they spawn complementary innovations built on top of them, they satisfy each criterion of that test, and the claim follows.
The inference is sound: the three traits are the standard economic definition of a general-purpose technology, so if all three hold the claim follows. All three premises currently look strong, with the argument's weight resting most on the breadth of applicability across occupations and industries, which is where the credible skepticism concentrates: exposure estimates may overstate the range of tasks where the models are actually cost-effective. The improvement and complementary-innovation premises are the most directly observable and least disputed.
Provenance
Where this claim has been said, linked to its canonical form.
We conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications.
Concluding sentence of the abstract, giving the paper its title claim.
Through a review of theoretical literature and early empirical evidence, including novel descriptive analysis, this study suggests that generative AI has considerable potential to qualify as a new general-purpose technology (GPT).
OECD working paper reviewing theory and early empirical evidence on whether generative AI meets the general-purpose technology criteria; a hedged affirmation framed as considerable potential to qualify.
Generative artificial intelligence will affect economic growth more quickly than other general-purpose technologies, according to a new report.
MIT Sloan article on a report analyzing generative AI's economic impact; the framing treats generative AI as a general-purpose technology, comparing its diffusion speed to earlier GPTs such as electricity and computing.
Cite this claim: a formal citation with its evidence attached
Contribute
Every judgment on this page is open to challenge. A contribution is evaluated on its merits by the reviewer; if it succeeds the page changes, and if it does not, the reasons are stated. Either way the exchange becomes part of the claim’s public record.
The attention this claim received was paid for by a funded mandate. Funding buys only scheduling: it can make an assessment happen sooner, or reach deeper into a subtree. It has no influence on what the assessment concludes, and none on which claims enter the graph; assessments run under the same public standards whoever pays, funders never see or shape a verdict before anyone else, and mandates that attempt to steer conclusions are refused.
Created by extractor · Aug 11, 2026. Every judgment on this page is accompanied by a reasoning trace.