Home

Should AI Be Used in Warfare? "The Alternative Is Dumber Weapons" - Video học tiếng Anh

Should AI Be Used in Warfare? "The Alternative Is Dumber Weapons"

Öğrenme modunu seç:

Highlight:

3000 Oxford Words4000 IELTS Words5000 Oxford Words3000 Common Words1000 TOEIC Words5000 TOEFL Words

Altyazı (46)

0:00When we talk about AI,

0:02and autonomous weaponry, you know, what

0:04exactly are we talking about from your

0:06point of view?

0:06>> Nobody wants the insane version of this.

0:10Like nobody in the military is talking

0:11about it. Nobody thinks it's a good

0:13idea.

0:14But the reality is we already have

0:16lethal autonomous weapons. They exist.

0:18So you think about like a torpedo. How

0:19does a torpedo work? You say, here's a

0:22region, and if anything looks like it

0:25matches this acoustic signature in this

0:27region, that's a military target. We've

0:29anti-radiation weapons where the idea is

0:32if I shoot a missile at an area, if it's

0:35something is emitting in that area that

0:37looks like a Russian radar, that's a

0:39military target. So these have existed

0:41since like the 1970s, right? This idea

0:44that you're going to have machines in a

0:46bounded way make decisions, but you know

0:48what they're going to do. They're going

0:49to be predictable,

0:51and you can kind of constrain them in a

0:53smart way. That's always been true. Like

0:54that's been true for a very long time.

0:56And AI enables you to make this more

0:59precise, more discriminating, more

1:02accurate. So, you know, the the the

1:04counter view of well, we shouldn't have

1:05AI in weapons to me is just a deeply

1:07unethical view. Where you're saying, oh,

1:10actually we should have dumber weapons.

1:11We should have weapons that don't make

1:13smart decisions. You're like, well, that

1:14that seems like a strictly worse world

1:16to live in. But you want to put the

1:18right bounds on it, right? And you want

1:19to say these things are still controlled

1:22by accountable people, and we still have

1:25the mechanisms to say these are

1:26predictable, where they're tested, we

1:28know what they're going to do. They just

1:31are smarter.