Unrestricted autonomous AI weapons are uniquely dangerous because they remove the human backsop for democracy and create untraceable, cheaply-executed attacks that are extremely difficult to defend against.

131:51London. He's he loves London and so he will never leave >> and so that I think is the they they they've been in planning for a while now. So Cor moved first, got got settled into the headquarters, etc., etc. So this is not this is longterm checked out for a while. >> Yeah. >> So this has been I think really useful conversation. I hope that people hear

132:17and you know we'll certainly put our clipping agent on the task of making sure that people hear your call to feel the AGI so to speak. you know, take really take uh the transformative power of the technology that they're developing into account as they look around their environments in these frontier companies and think about whether or not what they are seeing is okay or whether you know something maybe

132:43urgently needs to be done about it. Um I think that's really >> Yeah, >> please. >> And they they can I think I think they can go and receive counsel from the AI whistleblower institute. You don't need to decide, I'm going to go whistleblow. You can tell them, I think, under, you know, attorney client privilege, uh, what you're seeing and receive advice on what you should do about it. So I I

133:06would recommend to to anyone who's wondering that they should do that >> in So I I appreciate your example and again I appreciate that shout out and I I hope that that resource um you know is is valuable to people who find themselves in these uh I I have to imagine you know extremely stressful situations in many cases. >> Yeah. >> Um maybe in in closing I know we've

133:31already kept you longer than we booked you. Um tell us what you're doing now. you've, you know, had a little bit of time to decompress. You're starting up some new projects. What's the future for you personally? What, what contribution to this mega challenge do you aim to make uh in your next tour of duty? >> So, I've actually been basically just compressing or staying at a steady state

133:53since since putting putting the post out. Uh, and I've been I've been compressing. I've been working on this uh glove box tool, this AI sandbox tool. I started working on it in May before the giant blow up. Um I received some skepticism then uh about well we can keep using auto mode for now. But you know I think it's time I think it's time we stop messing around with these

134:17unconstrained agents that have just full permissions on our computers especially as alignment researchers. So my hope is to help the field of alignment have a tool that they can use reliably, easily, minimal fuss, uh, and and use responsible practices, you know, firewalled, uh, audit logged inside a virtual machine. That's what I'm working on right now. I'm a visiting engineer at FAR. Um, and I'm enjoying. >> Where are we on that? Can I use it now?

134:47What's the timeline to when I can use it? And if there's any you should build a beta yet. Yeah. Yeah. >> Yeah. So, agent glovebox uh on GitHub. >> Yeah, feel free to try it out. >> Agent agent glovebox on GitHub. >> Uh agent hyphen glovebox. >> Uh agent hyphen glovebox on GitHub. >> Cool. Um >> yeah, I'll be uh we'll be paying for a formal security audit soon. Um, and you