4 ms·

how many times has Kurzweil been correct? Or better yet, what's his success rate?

• arthurcolle 4y ago
  Quick Google says 147 predictions of which 117 are true. I don't have a good source though, and I'm not sure how strict the criteria for "pass" / "fail" for individual predictions were I used to like Kurzweil but he kind of jumped the shark with his more goofy predictions
  • lanstin 4y ago
    I went to a talk he had in 2010 or so when he was predicting we would ascend to god like digitally uploaded minds by 2030.
    • slowmovintarget 4y ago
      Which is a very large leap, because he fails to factor in pressures that could cause societal collapse, where the technology would be pushed far enough out of reach that nobody would be able to use it. Digital minds have an off-switch, and digital have-nots could still have axes and guillotines. But that's beside the point that encoding something has complex as a single human neural system requires several orders of magnitude more computing power than we've yet deployed in totality. Just the "parameters" of the human neural system would be in the 100s of trillions. A single neuron is more complex than most entire neural nets we've built. We'd have a long way to go, and that's just assuming everything about consciousness and awareness is strictly mechanism.
  • xref 4y ago
    > Of the 147 predictions, Kurzweil claimed that 115 were "entirely correct", 12 were "essentially correct", 17 were "partially correct", and only 3 were "wrong". Combining the "entirely" and "essentially" correct, Kurzweil's claimed accuracy rate comes to 86%. Lol I mean, those are the numbers Ray himself has decreed correct…