The biggest problem with AI

The issue is that everyone is treating artificial intelligence systems as if they are sentient, they are not and never will be, so we must stop anthropomorphising them. Asking AI to perform a task is not like asking a person. AI does not reason, it has no empathy, and cannot come up with an original thought.

It can be made to seem like it has all of these attributes, and every other human trait, but this is only if this has been coded into it. The code can be made to learn appropriate responses, but they will only ever be artificial (after all, the clue is right there in the name). So think of it like a Parrot who learns to talk, it sounds genuine but it is just a load of meaningless sounds to the Parrot.

So we have to be very careful of what we ask technology to do, because it will do exactly what we ask it to do.

Here’s an analogy: An inventor creates the perfect acting robot with all the latest tech to enable it to simulate every actor that has ever lived. It’s so good that people forget it’s a machine and start talking to it as if it’s alive. It’s about to perform its greatest ever role. Just before it goes on stage the creator says, “break a leg” to which it replies, “whose leg should I break?”

Post Views : 257