AI vs. Humans: A test on the strengths and weaknesses of AI
Applied Technology
Hugh Weirich
- Grade:
- 10
- Teacher:
- Jomaly Cotto
This project compared multiple AI models and a group of human participants on the same set of questions to explore their relative strengths and weaknesses. Responses were collected and scored, revealing different error patterns: AIs tended to make mistakes from misinterpreting questions or faulty reasoning, while humans more often struggled with time‑ or effort‑intensive items. Although individual humans performed worse than the AIs, a simulated combination of the best human answers suggested that collaboration could match AI performance. The study has limitations (small sample, limited documentation, and an untested team condition) and points to follow‑up work on actual teamwork and human–AI hybrid approaches.