There appears to be a little leaning toward "The Sky is Falling" in saying AI can be dangerous, but not specifying how it can be dangerous. Certainly a pile of code that can look up the amassed knowledge of mankind, can't make a nuclear weapon. It can't pick anything up. Of course it could guide people to pick up needed ingredients. But that can only occur when the person or persons involved asked how to make a nuclear or biological weapon. But of course, whatever AI can look up any search program can ultimately find. People can use AI to accelerate their gaining of knowledge. That is AI's primary function. AI will mature, just like dynamite production matured. But people had to put the physically destructive materials together. Evil, and stupidity, are human qualities, not programming goals. Ultimately, scientific knowledge on destroying anything, corrupting anything is in the public domain. AI cannot be limited to pick and chose what can be used to inform people about how to destroy people. Scientific information is not hidden, it is actively sought to make progress. I see no reason to try and hide scientific knowledge. It's in the public domain!!
No comments:
Post a Comment