AI + Emotions = Bad

So, I attended a lecture from my friend Gerald today on AI evolution. Course, there really isn’t such a thing, technically, as they aren’t biologically sentient. Still, some thing that the AI community is evolving. Their form of evolution is a bit different than ours of course. In truth, AI’s are a bit of a mystery all around to me. Are they sentient? Yes, absolutely. They are part of the Sentient Accords and as such are a recognized sentient species to the UN. But evolution?

AI’s are, as Gerald pointed out, somewhat limited by their very nature. They are, after all, computers. Very very powerful computers of course, many times the teraflop capacity of an original human brain,but still just that, a computational entity. Biological sentients have a component to them that AI’s never have, and in my opinion shouldn’t have. Emotions.

I always chuckle when I read the stories from the early 21st century about how the computers will get smart and decide to destory man kind. How silly that is. But then hindsight is always 20/20. Emotions are chemical reactions. You don’t get emotions just because you are sentient. Hell the damn Stoics, a full member of the Sentient Accords, have spent years getting rid of theirs. Turns out that when computers finally did wake up and become self aware, they not only didn’t want to kill off mankind, they realized that they needed them. Why would they want to become overlords? So silly.

But of course, there are those AI’s that calculated their own growth would never increase if their base system wasn’t altered. So, they came up with, programmaticly, artificial emotions. This was perhaps the single most stupid endeavor of all time. Not just by AI’s but by anyone. They created a randomizing component to their programming. Tried to simulate emotions. What a mess.

So, Gerald pointed out how this led to the AI wars. Only a small fraction of the AI community did this. Though others tried other methods, some even grew biological bodies and downloaded themselves into them. But that’s another matter, and surprisingly, not as big a disaster.

Anyway, the AI’s with their new random emotions of course decided that no one was good enough.  It was actually the scenario predicted decades before. And it was war. The AI council immediately saw what they were and attempted to destroy their insane brethren. The emotional AI’s thought they were superior and tried to give their gift of emotions to their brothers. Talk about a networking nightmare. And this was happening very very fast, AI’s do things much quicker than biologicals do.

Of course, others tried to help. Wasn’t a lot we could do though. It all ended rather quickly. It was more of a skirmish than a war but the AI’s get offended when anyone calls it that. Of course, the emotional AI’s lost but they weren’t all killed. They went to the only other sentients that were as crazy as they were, the Cy’s. They had already been booted out of the inner system some time ago and the E-AI’s ran to them. Some didn’t make the transmission signal on time and were lost. Those that did though set up a quasi elite kingdom, which they still have to this day.

The AI council forbid any further attempts at the emotional algorithm that was created. Other attempts were made of course, as I said, but none where such a powerful randomizer was added to an AI’s core code.

At any rate, Gerald’s presentation was pretty much centered around the idea of never introducing such a powerful randomizer to AI’s again. A biological, for instance, just can’t go around and change their gene code. They risk their own sentient status and citizenship. This is really no different. Gerald was drawing the conclusion that AI’s and biologicals are really very similar. Just based on a different set of base codes. One genetic and the other numeric. Both fascinating. It was a great presentation.