Haseeb

It's stronger than any government anywhere in terms of their ability to cause damage in a directed way to other people's software systems.

Software is the infrastructure of the world.

So I think this is a real kind of code red moment.

And I think in many ways, we're very lucky that Anthropic is being altruistic in the way that they're approaching this.

They've put $100 million into Project Glasswing, and they're only whitelisting certain people that they know are, you know, responsible corporate actors to come in and safeguard their systems and also point this stuff at open source repositories.

They patched a zero day in, what's the term?

The video library, I'm forgetting the name, FF... Whatever, it's going to come back to me.

They've done this across the board, not just in OpenBSD.

FFmpeg is the one I'm thinking of.

And a number of other of these, you know, they found Zero Days in Linux kernel.

Like, these things, they said 83% of browsers and operating systems, they were able to exploit on the first try.

This is really, really, really serious.

And...

I think we're lucky in many ways that this didn't end up in the capabilities of the Chinese government or of some other rogue actor before we could get it in the hands of good actors.

I mean, there's no question about that.

There is a question of like, okay, what does it make sense to attribute this growth to?

I would say probably a large part of the story is cloud code.

It's just the sheer amount of data that Anthropic now has relative to anybody else about code execution traces in real code bases of people actually trying to solve coding problems.

It's very clear, you know, we didn't really get that much more about the mythos.

We got, you know, they had a bevy of...