Quantifying Centuries of Tech Innovation by Filtering Wikipedia with LLMs
soumitrashukla9 · x · 2026-07-21
A recent NBER study demonstrates an innovative application of LLMs in economics and historical research. Researchers utilized LLMs to filter Wikipedia articles and extract descriptions of significant technological innovations over the centuries, subsequently pulling data on the time, place, and domain of invention to construct a massive historical dataset.
Empirical analysis based on this dataset reveals the evolution of innovations per capita across countries from 1500 to 1999. Notable findings include a steady increase in the US since its founding, while innovation intensities for Germany and France declined post-WW2. The paper also presents a model of innovation, knowledge flows, and technology adoption to explain the global distribution of innovation.
More from Research
- Project APE finds verifier reliability drops when papers contain multiple errors — soumitrashukla9 · 2026-07-22
- Project APE says verifier costs fell about 90x in a year as Chinese open models lead — soumitrashukla9 · 2026-07-22
- OpenAI-linked paper says capability RL can make models more reward-seeking — MariusHobbhahn · 2026-07-22
- Project APE builds its verifier benchmark from 100 AI-written papers with injected errors — soumitrashukla9 · 2026-07-22
- Paper proposes a CRED taxonomy and benchmark to measure research-error detectors — soumitrashukla9 · 2026-07-22
- OpenAI says long-horizon models need safety and alignment checks across full action sequences — rhiever · 2026-07-22