How AI Agents Are Becoming Part of the AEC Workforce

When you do a search in synthesis it for people that haven't seen it and they're listening, they can just imagine kind of like If Google and ChatGPT had a baby in a way, right? Google in that it's searching all of your internal knowledge and then chatgpt as it feels kind of like a chat interface. And yes, you can ask follow up questions and see if you ask, you know, what projects have we done with Atriums in Massachusetts and you know, our kind of company history, you can ask that question like, oh, shoot, and I actually want as a follow up, like projects that are over a hundred thousand square feet or whatever it is, so you can have this conversation with your firm's knowledge. One of the things you just said about like the friendly, the tone, that kind of piece, you can add that in as well. But one of the things, and maybe we'll get there later in the podcast, is we're starting to work on agents. And that basically lets you save some of those preferences, right? So I don't want to have to tell you every single time I want bulleted list, you know, I want you to use my words as much as possible. You know, please write it in this tone of voice. So over time, like we're making the technology really kind of easier to adapt to the way that you want to interact with it, if that makes sense. It's like, listen, software, I want you to love me the way I want to be loved, right? Yes, that's right. I don't want to have to tell you this. I mean, imagine, I mean, it's not that different. Well, let me take it. I'll go even bigger. Like one of the things that we've been really paying attention to and believe is true is that AEC firms will start thinking about their workforce as a hybrid workforce made up of employees and agents. And these AI agents you will want to think about like employees in many ways. Now they're not the same thing for a lot of, in a lot of ways, but it's true that they'll need learning and development, right? If you want agents to be able to answer questions to help people, they need to have knowledge in order to answer those questions, right? They need to have context in how you know your firm operates. And so one of the things that's interesting is if you had an intern, you would say like, hey, when I give you this assignment, you know, this is what we expect. It should do A, B and C. And the idea with agents is it would be the same way. I don't want to have to tell an AI system a hundred times. I wouldn't tolerate that from a person, you know, and. And so we shouldn't tolerate it from our software either is kind of the point of view. Or maybe said in a more positive way that sounded more negative than I meant it to come out. Like. One of the great things about humans is you generally don't have to keep doing that over and over again. Right. They learn very quickly and they pick up context. And I think that's one of the kind of potential benefits for agents is the AI tools of the future. You won't spend as much time handholding and repeating the same things over and over again, too.