Comments by "Martinit0" (@Martinit0) on "François Chollet: Keras, Deep Learning, and the Progress of AI | Lex Fridman Podcast #38" video.

  1. Well, the first problem as a researcher is to select a worthy problem to solve (assuming you are not working for a corporation that will hand you a problem to solve). This is not easy at all, if you want to be both useful and realistic. I agree with Musk here: the difficult part is not the correct answer but to ask the right question.
    1
  2. It's also incredibly difficult for a single human to make significant progress for humanity. We are just adding our little mosaic piece (if we are smart and motivated enough). The very best of us may be able to add several pieces (think Einstein) - but we all stand on the shoulders of giants, our ancestors who added their pieces. All that only works out so well, because we can effectively communicate across space and time using written language.
    1