on that is that was a terrible idea. And, and, and, and let me give you some very specific examples on that. Uh, one example, absolutely wonderful example that Ori, uh, that's a, a company that I believe we've, uh, we chatted with before that company did an experiment where they took an AI code developer and cranked it up to maximum, gave it, you know, full access to an entire simulated, uh, development and production environment and told it, okay, go develop this app. And then in the process, they then said, okay, Hey, we need to put a code freeze in place. So they tell the, the AI developer, Hey, we've got a code freeze in place. Don't make any code changes. And it ignores the code freeze, tries to do something, panics because it can't make the modifications. So instead of modifying the code, it then deletes the entire production database. Obviously not a good thing. Uh, a more recent specific example, uh, that actually happened at meta is one of their, let's see, it's their AI alignment director was experimenting with a AI called open claw. And for those not familiar, open claw is an open source AI agent that can basically function as a virtual employee. Like it can take actions on its own and do whatever it feels is appropriate. Well, apparently at some point she must've asked the AI to go clean up her inbox and her email. So it did exactly that. It started to delete every single email in her inbox. And despite her repeated, uh, asking it to stop, it ignored her telling her to stop and completely wiped out everything in her email. So I think we're, we're finally starting to learn. And unfortunately, we're learning the hard way that maybe just implementing this AI stuff, especially the agentic AI wasn't such a good idea when we don't have the proper security controls in place. So finally CISOs are, are, are, are finally getting involved and they're, they're basically saying, yeah, you know, we really need to secure this stuff before we deploy it. Again, very hard for me to disagree with that, right? When you start taking and talking about any tool. And again, I will always take the perspective that AI is yet another tool in your toolbox in order to do things. Okay. So when you are looking at any tool, you should look and evaluate what are the risks associated? What are the costs associated? What is the brain