AI Tests Invention Timelines

Updated: 2026.05.12 2H ago 1 sources
Large language models can be used to systematically ask historical counterfactuals — e.g., could this technology have been invented earlier — and thereby quantify when inventions became feasible versus merely culturally 'ahead of their time.' Using an LLM to scan prerequisites and dates produces a replicable, semi‑quantitative map of technological possibility. — If LLMs can robustly assess when inventions were actually feasible, historians, policymakers and innovation strategists can replace anecdotal 'ahead‑of‑their‑time' narratives with evidence about prerequisites, altering how we reward, fund, and teach innovation.

Sources

Ideas Behind Their Time: Part Two
Alex Tabarrok 2026.05.12 100% relevant
The article cites Brian Potter using Claude (an LLM) to examine many inventions and concluding most could not have been invented earlier; airplane example (engine power circa 1880) is given as concrete evidence.
← Back to all ideas