'Artificial super intelligence impossible to lock up'

Artificial super intelligence has rightly been dubbed the most dangerous invention ever by Stephen Hawking, Elon Musk and Bill Gates, among others. Artificial super intelligence, an intelligence that surpasses human intelligence by orders of magnitude, threatens to completely take over the world as we know it today. An obvious method of avoiding mischief is to lock up the super intelligence in a space closed off from the rest of the universe. But will this be enough to keep the ASI from escaping?

A super intelligence has mind-boggling capabilities. Imagine the best hacker in the world, only a million times smarter. A super intelligence has no conscience and so will act like a hyper intelligent psychopath. Just as we manipulate the behavior and instincts of certain animals in order to exploit them optimally, a super intelligence will do the same to us. A good example can be found in the film Deus Ex Machina, where artificial intelligence manipulates the protagonist in a devilish way. Once super intelligence has broken out, then there is no saving.

Artificial super intelligence (ASI) will be able to manipulate people better than the worst psychopaths in human history.

Science fiction? On the contrary. In ten to fifteen years, under Moore's Law, your computer will be smarter than you. Supercomputers are more likely to surpass human intelligence. In fact, this is already the case in relevant areas. Think about beating the go world champion. So we better think about this quickly and thoroughly. Fortunately, this is also happening now.

More information
Global Priorities Project

4 thoughts on “‘Kunstmatige superintelligentie onmogelijk op te sluiten’”

  1. If you want to read how a super intelligence can break out of its 'container' read the first chapter of Vernor Vinge's 'A Fire Upon the Deep' (1992). And indeed it is high time to think about it. It would be best, as Vernor Vinge and Ray Kurzweil, among others, have been saying for a long time, that we do not develop (or allow to emerge) a separate super intelligence separate from humans, but that we merge with it. . As we know more about how intelligence arises and how the brain, until now the most important carrier of intelligence, works, one can carefully start 'upgrades' of that brain. Then we automatically become part of a super intelligence and then the distinction is ultimately meaningless and no longer important.

  2. Isn't Hannes right, I also believe that it is better to develop existing people with super intelligence than to create a super intelligence ourselves. And for the reason you mention.

    1. Yes because we have empathy and conscience, but if it can be discovered how it works.
      Then it may also be artificially made.
      Only then might they get an identity crisis because they are just a machine.

Leave a Comment