Computer scientist and cognitive psychologist Geoffrey Hinton says despite its potential for good, AI could one day escape our control.

Godfather of AI tells ‘60 Minutes’ he fears the technology could one day take over humanity::Computer scientist and cognitive psychologist Geoffrey Hinton says despite its potential for good, AI could one day escape our control.

  • RonSijm@programming.devEnglish
    95·
    1 year ago

    What we have is machine learning, just an algorithm that takes input and gives you output. It can’t act on its own.

    Isn’t that basically what “real learning” is as well? Basically you’re born as a baby, and you take input, and eventually you can replicate it, and eventually you can “talk” for example?

    But in the training data something was off, suddenly your AI is racist and gives every black person a lesser amount.

    Same here, how is that different from “real learning”? You’re born into a racist family, in a racist village where everyone is racist. What is the end-result; you’re probably somewhat racist due to racist input - until you might unlearn that, if you’re exposed to other data that proves your racist ideas were wrong

    If a human brain is basically a big learning computer, why wouldn’t AI eventually reach singularity and emulate a brain and beyond? All the examples you mentioned of what it can’t do, is just stuff it can’t do yet

    • dgmib@lemmy.worldEnglish
      62·
      1 year ago

      All the AI we have today is, at its core, just pattern recognition.

      ChatGPT can answer questions because it’s been shown a VERY large list of questions and their right answers. ChatGPT has no idea what the question is or what the answer means. It just has an algorithm that knows that a particular answer fits the pattern of “a correct answer” for that question better than any other answer.

      It can’t “reason“ or “think” in any way. It’s not going to become self aware or set its own objectives. And so far we don’t have anything close to true general AI, we don’t even know if it’s possible.

      There’re still risks from the current AI though. AI will sometimes find unanticipated and undesirable solutions that technically meet the goal it was given. A “Terminator” style future is unlikely without artificial general intelligence, but it’s not completely unreasonable to think of a scenario like “I, Robot” where a “dumb” AI subjugates humanity as a solution to a more altruistic goal like ending war or famine, because it’s a solution that matches the pattern it was told to look for.