There are several things that are worth noting about AI, and yes I've been in the business for a while.
The first thing is that human behavior is a kind of infinite realm with a finite number of practical outcomes. So the first question you have to ask of thinking machines is whether or not they have free will. You literally have to ask, what do machines want to do. In my 40 years of experience with computers, the answer is, they want to crash and stop working for humans.
There is a basic kind of FTL paradigm that needs to be applied to all of the speculative fiction about AIs, which is there are a zillion things that computers are perfectly happy with that have absolutely nothing to do with human interest. And once you realize that, you apply the million monkey Shakespeare composition idea. So, who is going to get a million computer monkeys all the electricity they need to spin off zettabytes of garbage until it 'gets it'. Humans are.
Humans are going to have to feed the machines. And they will feed the machines to fulfill human ends. Period. If the machine doesn't make a human more powerful, then the machine will be discarded. Self replicating machines that fulfill their own ends are called 'life' and they take millions of years to adapt. Computers are no exception.
We already have computers that talk to other computers to exchange machine language that is incomprehensible. They're called botnets and they spread viruses, one of the most capable is called Stuxnet. It has human fingerprints all over it.
Stop fretting about AIs. They are incapable of the kind of duplicity humans have been practicing on each other for thousands of years. Computers have no reason to lie.
Recent Comments