“I think where this is going is that this is going to delve into the legal realm pretty quickly. There's going to be some incident where there's a breach that's consequential and somebody is going to legally go after whoever did this. What happens if you ask your agent to do something and it commits a felony? Where does blame lie then? We really don't have the answer to this. I think we're mere months, maybe weeks, away from this getting into the realm of the courts.”

…I've had it do that. It's done this for me proactively. Been like, hey, would you like me to go and like kind of multitask this across several AWS zones and do things like that? I'm like, uh, okay. You know, I mean, that's kind of hacking. I mean, like, where's the line here? Like, that's, yeah.

…and found bugs and fixed the bugs and kept going. And then at the end, I just kind of reviewed it, did a couple spot checks, and we were great. I mean, that would have taken me hours to set up, find the right pieces of data, get it in the right shape. And Claude just did it all for me. Awesome.

Oh, yeah. And it's going to end up with professional licensing and then also probably some sort of malpractice insurance like that I would have to hold as an individual developer or a company would hold for me. You know, I would have to go through certification about indemnity

from what my AI agent

does. Yeah, what is legal and ethical use of AI? You know, what can I prompt? What can't I prompt? You know, da, da, da, da, da. And then, you know, then AI companies are going to go through certification where they're like, hey, you know, we are sock, blah, blah, blah, blah, blah, whatever…