George Butler

Typically, click all the traffic lights, type some distorted letters, find the bicycles, tick a box, it's to basically prove that you're a human and not a robot.

Because typically, up until now, it's been very hard for robots to do that stuff.

The problem is that AI has got very, very good at doing these things.

So this is like just a core layer that the internet runs on.

You know, whether, like you said, you're buying tickets or anything like that, it just stops bots going in and frauding stuff.

The new AI systems have got very, very good.

They can mimic human behavior.

They actually can browse the internet like a human.

So if it can watch humans doing it, it can do it itself.

So you've got this whole thing on the internet that we've sort of built to stop fraudsters or stop bad actors doing things.

But we might very quickly have to start thinking about this again.

And I was reading lots of forums around where this ends up.

You know, effectively, you get a lot of the internet online

which is AI talking to AI.

You get sort of a fake review with lots of fake comments under it.

You know, you get this sort of dystopia around what it could be.

What we'll need to do is to figure out a way of verifying humans in an entirely different way.

Now, of course, that can be things like we've got face ID on their phone.

It could be that or touch ID, things like fingerprints.

It could scan your retina.