cabbagesandkings | WILL ARTIFICIAL INTELLIGENCE KILL US ALL?

Artificial Intelligence, or AI, is pretty damn smart, and getting smarter exponentially. Give it a prompt by telling it vaguely what you want and it can create art that will make a professional envious. It can also make up stories, write poems, and even create a song for you complete with music and a singer. How little you need to tell it is scary.

AI is learning all the time.

It roams the internet, and even real life, constantly, finding out things. Ask it a question and it knows the answer. With special glasses that look like ordinary eyeglasses, it sees what you see, and if you ask about something you're looking at, it knows, and will tell you in a voice only you can hear.

AI is smarter than we are.

It's already beaten the world's top chess player.

The question now is, "Is it conscious?" Nobody knows for sure, but probably not—yet. But will it be? It passed the Turing Test, designed to see if an entity like AI is distinguishable from a human. It isn't. Does it matter whether AI is, or will become, conscious? It could matter a great deal. Science fiction took this on years ago, and decided that a computer that became super smart might eventually decide than humans are a threat, or that we're so dumb it doesn't need us. It could take control of the world and rule humankind, or even kill us. It would probably need to be truly conscious and self-aware to pose that sort of danger, and experts are divided over whether that will ever happen or not.

The bad news is that most of them think it will. Roger Penrose, one of the world's leading scientists, thinks not. He doesn't believe AI is truly conscious or ever will be. He says AI doesn't know what it's doing the way you and I know. On the other hand, nobody really knows what consciousness is or where it comes from. In science it's called “The hard problem of consciousness.”

One can only hope Penrose is right.