>Yudkowsky's argument that a superintelligent AI will inevitably destroy humanity seems to have no flaw. From trying to see what his argument is, it doesn't se

>Yudkowsky's argument that a superintelligent AI will inevitably destroy humanity seems to have no flaw. From trying to see what his argument is, it doesn't seem to have much rigor either, basically that AI smarter than us may figure it doesn't want us. Which doesn't prove it'll happen. Even humans who generally are not that nice tend to try to preserve species and I expect we'll try to give AI better values than us warlike primates.