AI Startup Assistant Effectiveness Hinges on Execution
daveholtz · x · 2026-07-11
This field experiment, published in *Management Science*, investigated the impact of generative AI on entrepreneurial performance using a sample of **640 entrepreneurs** in Kenya. Roughly half of the participants received a **GPT-4-powered business assistant** via WhatsApp, while the other half received traditional business training materials. The results show that **on average**, access to the AI assistant did not lead to a statistically significant increase in revenue or profit, but this average masks substantial heterogeneity. **High-performing entrepreneurs** were more likely to translate AI advice into **specific, contextualized improvements**; **lower-performing entrepreneurs** more frequently adopted **generalized strategies** like offering discounts or running ads. The study further found that the broad topics of questions asked by both groups were similar, and the AI's advice was largely comparable, suggesting the disparity stems from **how users filter and execute advice**, rather than the prompts or the AI's output. The authors conclude that in open-ended tasks, AI's effectiveness depends on whether human judgment complements it.
More from AGI Musings
- Larry Fink says China is ahead in the AI energy race, citing 100 GW nuclear buildout — rohanpaul_ai · 2026-07-21
- Jamie Dimon says bureaucracy, not AI, is the real system crushing intelligence — r0ck3t23 · 2026-07-21
- OpenAI and Anthropic’s internal models are said to be far stronger than today’s public systems — scaling01 · 2026-07-21
- Superintelligence and robot abundance will force a new social contract — Dr_Singularity · 2026-07-21
- The Guardian examines how AI companionship is turning intimacy into an economy — nordicinst · 2026-07-21
- A frustrated user says modern AI keeps hallucinating on real-world repair tasks — doochenutz · 2026-07-21