Comment by morby

I can see people using AI for setting up simple solutions. But AI, as it currently is, is dependent on human innovation to be able to generate results. Any novel problem will need human minds. On top of this AI is currently horrendous at coding things with complexity. Maybe some higher complexity solutions can get done with high cognitive tax in human prompt input, but there are serious diminishing returns on that. AI still needs major breakthroughs to get to that point. And when it does it might very well have outpaced humans in a multitude of academic fields. Including the ones listed by Jensen Huang (the sciences).