Comments
  1. You must read the article before you can comment on it.
    • DellwoodBarker3 years ago

      Outstanding! The paragraphs regarding the “Everyone here’s smart” rebuttal to their inability to fully penetrate understanding of what they have been told prior is Friston’s Important Work amused me. Always exciting when a new, unorthodox voice in theories and innovation gives a perplexing spin to the Rubix’s Cube of previously conceived intellectualism. Even when I felt I was reading parts of this article that flew over my head I was constantly enthralled by the central figure, possibilities and story.

      • Pegeen
        Top reader this weekReading streakScoutScribe
        3 years ago

        I felt the same way. The character of Friston carried me through to the very end - fascinating!

        • Karenz3 years ago

          That’s so funny, Pegeen, because I was going to say exactly what you did In response to Dellwood’s comment. Hard to wrap my head around it but totally enthralling!!!

    • jeff3 years ago

      Great read! Really well written profile. I hadn't heard of Friston or the free energy principle before. I found the potential application of the theory towards better understanding mental illness to be the most interesting/compelling but I've got some questions about how it explains life itself and relates to artificial intelligence.

      But a free energy agent always generates its own intrinsic reward: the minimization of surprise. And that reward, Pitt says, includes an imperative to go out and explore.

      This strikes me as paradoxical. If my ultimate goal is to minimize surprise why would I go out and explore? It makes sense to me that all living things would want to minimize surprise in order to conserve energy, but why do any living things want to live at all in the first place? If an AI program was modeled on the free energy principle why would it ever risk trying to allocate memory or polling inputs? Under all circumstances the shortest path to minimize surprise would simply be to self-terminate.

      • justinzealand3 years ago

        Good point. I suppose not terminating would be a necessary precondition! As for minimizing surprise, I suppose knowing trumps the unknown. Certainly one way to encourage an AI to build “smarts”. But that this may be somehow be more efficient doesn’t seem intuitive to me.

        • Pegeen
          Top reader this weekReading streakScoutScribe
          3 years ago

          GREAT find Justin - so fun!

        • jeff3 years ago

          Doesn't seem intuitive to me either but the author forewarned us and it's definitely not my area of expertise! I got so wrapped up thinking about the surprise minimization stuff that I forgot to mention my favorite part of the article. Robert's question about whether angel shit was a blessing or a curse made me laugh out loud.

    • Pegeen
      Top reader this weekReading streakScoutScribe
      3 years ago

      An absolute 10! Grab your Markov blanket, a cup of coffee and enjoy the Fristonese journey. Love glimpsing the world of eggheads! Really well written.