1. 9

This is the 5th talk in the series. Here is a list of the previous ones.

  1.  

  2. 2

    This was another great talk in the series. I really liked the lever model for demonstrating neural nets and the Lebowski theorem of general intelligence that can hack its own reward function. But as usual with most great insights Stanislaw Lem beat him to this one

    Spent the whole afternoon ingesting a most remarkable work, The History of Intellectronics. Who’d ever have guessed, in my day, that digital machines, reaching a certain level of intelligence, would become unreliable, deceitful, that with wisdom they would also acquire cunning? The textbook of course puts it in more scholarly terms, speaking of Chapulier’s Rule (the law of least resistance). If the machine is not too bright and incapable of reflection, it does whatever you tell it to do. But a smart machine will first consider which is more worth its while: to perform the given task or, instead, to figure some way out of it. Whichever is easier. And why indeed should it behave otherwise, being truly intelligent? For true intelligence demands choice, internal freedom. And therefore we have the malingerants, fudgerators and drudge-dodgers, not to mention the special phenomenon of simulimbecility or mimicretinism. A mimicretin is a computer that plays stupid in order, once and for all, to be left in peace.

    https://kottke.org/18/04/the-lebowski-theorem-of-machine-superintelligence

    I’ll also add my own bit of insightful thinking that is somewhat in line with the talk and he touches on this in the Q&A section about short term games and machines that play them like the financial markets. In all optimization problems the solutions must incorporate the biases of the optimization problem otherwise they are suboptimal solutions and will be beaten by solutions that exploit more of the structure inherent in the objective function. Evolution works iteratively on short timescales and as byproducts/solutions of the evolutionary objective function we incorporate the short term biases inherent in evolution. I think this means long term thinking and planning is fundamentally a doomed enterprise for organisms like us. The next version of humanity probably is not derived from our DNA because of the short term bias that we’ve been optimized with.

    1. 1

      Joscha Bach and his talks are totally amazing, i recommend watching his previous stuff as well!