Felipe A. Csaszar, Harsh Ketkar and Hyunjin Kim. Strategy Science 9(4), 322–345 (2024).

https://doi.org/10.1287/stsc.2024.0190

Inspected author manuscript: https://arxiv.org/html/2408.08811v1, sections 3.1–3.2. In an experiment, 250 evaluators with investing experience rated versions of ten business plans; GPT-3.5-generated versions received higher average ratings than entrepreneur versions. In a separate analysis of 138 text-only competition plans, AI scores were positively correlated with judges’ scores. These are ratings and stated interest, not verified claims, funded investments or realized returns. The study motivates comparison with capable AI assistance; it does not test 1Cademy or establish an exponential increase in startup ideas.

Publisher identity and author manuscript inspected 10 November 2026.