Could an artificial-intelligence agent pass an introductory physics course?

Massive pre-trained language models have garnered attention and controversy due to their ability generate human-like responses: frequent indistinguishability from human-generated phraseology narratives, the fact that convincingly presented arguments facts are frequently simply false. Just how these responses when it comes dialogues about physics, in particular standard content of introductory physics courses? This study explores question by having ChatGTP, pre-eminent model 2023, work through representative assessment an actual calculus-based course grading same way human would be graded. As turns out, ChatGPT narrowly pass this while exhibiting many preconceptions errors a beginning learner.