>Musk outlined a hypothetical situation, for instance, in which an AI could pump up defense industry investments by using hacking and disinformation to trigger a war.
What the heck is he talking about? With my limited exposure to AI and neural networks, there really is no algorithm that can make algorithms. And therefore, AI doesn't really "think". Sure you can train a neural net to pick out the "diamonds" in a sea of garbage, but that is still not "thinking", merely going on an educated guess backed by statistics. Or am I missing something?
If he would be really concerned about AI, then there would be no auto pilot in Teslas. I'm pretty sure he would benefit from a regulated AI research somehow.
So that's why he is actually saying it. Think about the rest of the population. They have even less of a clue.
Obviously, AI is not right now in a state to be a thread. However, if you extrapolate - we are getting close to the singularity.
If you assign any probability greater than zero to that the singularity will happen in your lifetime, it is probably a good thing to start thinking about, how to deal with it.
A lot of AI apocalists are worried about the "singularity" which is the point at which an AI could program itself and become aware and have a largely negative affect on the world.
if you have an AI that is pretty smart, and you add a lot of money, then you have a pretty big problem as since you can scale it, and let it do what you want. i.e. side stepping democratic approaches. or causing large scale unemployment.
Later, you have the selfaware-ness problem, but that's less of a concern in the shorter term.
What the heck is he talking about? With my limited exposure to AI and neural networks, there really is no algorithm that can make algorithms. And therefore, AI doesn't really "think". Sure you can train a neural net to pick out the "diamonds" in a sea of garbage, but that is still not "thinking", merely going on an educated guess backed by statistics. Or am I missing something?