@dalias While the details are new, the pattern isn't. I used to review papers for conferences and we would regularly get people applying the general method in vogue for people who don't know much about a subject area to solve a problem. At one time it was genetic algorithms. Simulated annealing was hot before that. Then it was machine learning, and now we ask an LLM to spit out an answer. Anything other than learning a subject deeply and developing methods based on actual insight. I wasn't popular for point out things like this.