Researchers from Cambridge and the Hebrew University tested ChatGPT with a 2,400-year-old “doubling the square” puzzle to explore AI’s reasoning. While the chatbot solved the square challenge, it faltered with a rectangle, wrongly claiming no solution existed. Published in the International Journal of Mathematical Education in Science and Technology, the study suggests ChatGPT improvises like a student, sometimes creative but often flawed. Scholars warn its proofs aren’t always valid, stressing critical evaluation in education.
from Tech and Gadgets-Tech-Economic Times https://ift.tt/DeuVGRO
No comments:
Post a Comment