Examining the Potential and Pitfalls of ChatGPT in Science and Engineering Problem-Solving
This provides insights for educators and researchers on AI's limitations in STEM problem-solving, though it is incremental as it applies an existing method to new data.
The study tested ChatGPT's ability to solve physics problems, finding it solved 62.5% of well-specified problems but only 8.3% of under-specified ones, with failures due to modeling, assumption, and calculation errors.
The study explores the capabilities of OpenAI's ChatGPT in solving different types of physics problems. ChatGPT (with GPT-4) was queried to solve a total of 40 problems from a college-level engineering physics course. These problems ranged from well-specified problems, where all data required for solving the problem was provided, to under-specified, real-world problems where not all necessary data were given. Our findings show that ChatGPT could successfully solve 62.5% of the well-specified problems, but its accuracy drops to 8.3% for under-specified problems. Analysis of the model's incorrect solutions revealed three distinct failure modes: 1) failure to construct accurate models of the physical world, 2) failure to make reasonable assumptions about missing data, and 3) calculation errors. The study offers implications for how to leverage LLM-augmented instructional materials to enhance STEM education. The insights also contribute to the broader discourse on AI's strengths and limitations, serving both educators aiming to leverage the technology and researchers investigating human-AI collaboration frameworks for problem-solving and decision-making.