Retired U.S. Army General: ‘We absolutely will’ give control over lethal strike to A.I.

Retired U.S. Army General Stanley McChrystal joins 'Influencers with Andy Serwer' to share his biggest fears regarding artificial intelligence.

Video Transcript

ANDY SERWER: I want to ask you about AI, artificial intelligence, because you wrote, "ceding the ability to manage relationships to an algorithm, we rolled a dangerous die." What are the specific uses of AI that concern you and then we can talk about AI weapons and that's really scary stuff. But let's talk about it generally and then specifically with regard to the military.

STANLEY MCCHRYSTAL: Let's start by something we all get. We call company X and we get this recording that says if you're calling about so-and-so hit one. If you're calling about so-and-so hit two. And you go for a while and by the time you get to 8 and they didn't cover your problem, you're furious. And you just want to talk to someone. You want somebody to take your problem for you.

And if I'm a company and I put that in, I save money. I don't have to train anybody. I don't have to give them overtime or any of that. But what I sometimes don't know is how many times does Stan McChrystal or Andy Serwer recall and then I just say I'm done with you and I go to a competitor.

And so we need to understand that when we use technology, there can often be two edges to the sword. I think that AI ups the game there. We start to think that artificial intelligence can do processes for us, some of which you can do and some of which it will make decisions different from the way that we want them to be made. Or either impact us or the organization. And the problem is knowing it. It's hard to have a complete understanding of in a modern organization now, what decisions are actually being made algorithmically and which are being made by people.

And so when you don't have that, I would argue you have the risk of no longer having real understanding or control of your organizations. You mentioned AI controlled weapons. People say, we'll never give control over a lethal strike to artificial intelligence. And I say, that's wrong. We absolutely will. Because at a certain point, you can't respond fast enough unless you do that.

A hypervelocity missile, hypersonic missile coming at a United States aircraft carrier, you don't have time for individuals to do the tracking. You don't have time for senior leaders to be in the decision loop or you won't be able to engage the missile. So you've created technology. You put in processes for it to operate. But then to operate at the speed of war, you're essentially turning it on and trusting it. And that can be pretty frightening, particularly, if the potential of malfunction or spoofing or any of those other things are in.