So, After 1 year of computational ph.d., I really pushed a lot the AI models. But honesty, also if they can be really good, I think that it does not really spee

So, After 1 year of computational ph.d., I really pushed a lot the AI models. But honesty, also if they can be really good, I think that it does not really speedup a lot in long term a junior. You cannot really present non-validated results, and the validation of the huge amount of unorganized things that they generate is really stressing. So yes, for a while I was losing motivation, because I felt like I had to produce more but I could really follow that productivity. To really get motivation, u need to spend the correct amount of time of understand and make your hand dirty when it is necessary. I noticed that if I try before by myself, I understand what I am doing, and after I just optimize the code using an LMM, it is way more motivating