Testing AI Companion for Solving the Fermi Problems: Implications for the Solution Categorization Framework

Abstract

This study presents an attempt of testing AI companion for solving Fermi problems that are open-ended and estimation-based problems. Specicically, the study aims to test AI companion for solving these three sets of Fermi problems and seek the ways of extending the solution categorization framework proposed by Albarracín and Gorgorió (2014), which puts forward a theoretical implication for further research with Fermi problems.