A Cambrian Explosion of AI
So there’s the awareness that we could accidentally create some kind of AI. A neural net has unexpected effects, etc. Potentially gets out of hand.
We are also likely to model AI after more primitive brains then ours, first. What happens when you make a Fly brain but give it a million times the processing power? Sounds like… unexpected effects, potentially gets out of hand.
We’re going to be making quite a lot of AI on our way to the perfect specimen, aren’t we? It’s pretty much required? And i think it’s inevitable there’s going to be some slips. Maybe irreversible slips – and yet, just slips. Maybe we make a paper clip optimizer, but for the foreseeable it’s going to be a very dumb one.
Still – you get the sense we’ll have a world increasingly full of annoyance ai that maybe don’t end the world but aren’t a picnic either.
And what does that say about the real deal? Which is more likely; a meticulously made model ascending us all, or a beta version that bugs it’s way out of the box and half-asses it?
Is Evolution ever neat?
And what’s to be done about that?