AI has not yet reached the point where it can convincingly generate long presentations with the precision required for complex, technical academic topics. It wi

AI has not yet reached the point where it can convincingly generate long presentations with the precision required for complex, technical academic topics. It will eventually get there, at which point an additional layer of proof of work will need to be introduced. This is an adversarial race. I think this is an interesting and effective solution, at least for now. Similarly, graduate and master's theses should focus more on the presentation and on eliciting knowledge from the students through critical, thorough questioning than on the tangible outcome of the project, which can easily and bindly be obtained with AI these days.