Gigaom AI Minute – April 20

Transcript

When you type, "Is artificial intelligence..." into Google to see what it auto suggests as possible searches you may want to do, the number one suggestion you get is "Is artificial intelligence a threat?" Followed closely by, "Is artificial intelligence dangerous?" Why all of this fear around this technology, which to date hasn't harmed anyone? I think there are two reasons.

One, we have a long tradition of fearing the things that we create, that they're going to rise up and destroy us. You can see this in stories like Frankenstein. You can even see this further back in Greek plays, like Oedipus, that had children killing their parents.

The second source for it must be popular media and entertainment, which shows all kinds of scenarios where artificial intelligence goes berserk or decides to take over. We all know HAL in 2001, and all of the rest of the examples. Of course, the truth is that that does not in any way indicate what is going to happen. That's a logical fallacy known as generalizing from fictional evidence. Interestingly, one of the autofill suggestions is, "Is artificial intelligence capitalized?" Even on that there isn't widespread agreement, which just goes to show where we are in the science.