How much do you use your own experience to adjust or refine the code produced by AI as you develop Show Hub?

29:23engineer is you need to understand it. So you can't just get in a car without thinking about the engine and saying, oh, how does this all work? Oh, I hear a sound and it's sort of needing to, that need to know how everything's put together. So even though I might have AI do a bunch of the the code, I still need to understand it so I can understand the whole. So I definitely have some of

29:47that. But I don't, there's a temptation that if you handcraft something, you hold onto it longer than you should. And if there's a feature that's not working or you need to replace it, you should just be very comfortable throwing it out. And if you feel, if you have too much pride over that little piece of code because you did this little clever loop or whatever and that's going to slow

30:10you down and get in the way so i try not to do too much of that the and what i do in my day job is is we've got software that does this at a super macro level or or a larger level where where we're coordinating across many hundreds of people doing these really large IT projects, moving data centers and whatnot. So some of the learning of how do people work together is definitely coming into play there.

30:45One other thing that I'll add in, obviously AI made a big difference in enabling this to all come together really quickly. We are seeing AI, especially agents now, coming into different areas. And I'm seeing it at work. I'm seeing it everywhere where you have an agent do a bunch of tasks for you. That's baked into the new task system because we're going to get to the point where, you

31:14know, like Vic might be welcoming a guest and maybe he's got an agent that goes around and checks things. But he wants to know, did it kick off? off, did it finish, what the results are. So just like coordinating with other people, we're going to have to have systems that if we have automation and agents running around, we need to be able to coordinate with those so that it's not us and them, it's all one cohesive

31:38team. I was struck by what you said in the first part of your answer about killing things that you liked and getting to the point. As someone who writes a good little bit, I remember hearing one one of my editors saying, you know, you really have to do, you have to learn to kill your babies, that you have to learn to take something that you loved. And even if you loved it and you thought

32:00you executed this line of script or whatever particularly well, if you stand back from things and you find out it's, it was not the right way to go, you have to learn to give up on the things that you particularly liked and take them out for the, for the, to, to, to get closer to a higher higher level of better writing or whatever.

32:22And it sounds like you're saying the same thing, that part of your learning process here was even if you put in a feature and you thought people are going to love this, everybody's going to die for this little feature, it may not end up being what the whole thing needs, right? Right, right. And it's a little easier now with AI that if I'm doing it, I might spend minutes, you

32:43know, working with AI to develop some features. So if I don't like it, it's only a few minutes lost. If I handcrafted something, yeah, if I spent a week building something and it's like, we've got a better way to do it, you know, I'm going to be pretty hesitant and argumentative. I don't know. Yeah, no, I think that sounds very real. Let's go to our next question.

33:08Coming in from Pittsburgh, Pennsylvania, Josh Kaufman asks, how does feedback from the people using custom software shape its evolution? And how do you distinguish an individual preference from a change that benefits the whole organization? It's a fascinating question. And how do you, let me just spin it into this. When you have a big organization like ours here and you get a lot of feedback on things,