“People Make Poor Monitors for Computers”

That’s the title of a new book that I intend to try to read, time permitting. There’s sort of an encouraging thought in there: we’ve gotten so good at designing computers, that giving humans too much control over them may no longer be a good idea.

  • unbound

    Link is messed up. Should be:


    Overall, my main concern with the concept of not allowing human beings to take control is that the complexity of code keeps increasing, but time and funding for testing continues to be decreased to save on costs. So while the notion of human beings not being in a good position to take over in all situations has some serious merit, there is a growing problem on the horizon that speaks to the continuation of the need to take control…just for different reason than has been historically considered.

  • Pingback: Most useful Discount On AngelCare 3-in-1 Video Baby Monitor

  • piero

    In my opinion, it all depends on motivation, which is a very difficult concept to translate into programming code.

    Humans have been shaped over billions of years by a mindless process which has made survival our primary goal. An artificial intelligence might well have other goals, unless we are stupid enough to replicate in it all human motivations, and endowing it with a superhuman intelligence.

    Imagine the following scenario: biology, neuroscience and computer science manage, in a joint project, to generate a 3-foot scorpion with superhuman intelligence and a lethal venom for which no antidote is known. That project would be uniquely stupid.

    But in the realm of artificial machinery, I’m rather more optimistic. Unless a nutty scientist comes up with a way to obtain energy from an as-yet unkown source, any dangerous artificial intellligence can simply be unplugged.

    • sunsangnim

      Your scorpion idea is hilarious. I’m sure the military would order some.

  • Pingback: yellow october