Latest / Engineering Influence from ACEC / Unveiling AI's Rapid Evolution with Deltek
Transcript
- 0:00Music.
- 0:09We're coming to you live from our annual convention and legislative summit in Washington, D.C.
- 0:14I'm very pleased to be joined today by Brett Tusshaus, the vice president of
- 0:19product management from Dell Tech.
- 0:21Brett, welcome back to the show. Thank you for having me once again.
- 0:25Yes. Yeah. I think we were talking and this is the fourth time you've been on
- 0:27the show. It is the fourth time. Definitely.
- 0:29So, I mean, if we were to turn back the clock to the first one,
- 0:32second, third, fourth, I mean, and we're going to talk about AI, right?
- 0:35That's really why we're here lightning's lightning
- 0:38fast speed is that a good way to describe the changes that are happening on
- 0:41the i don't i don't think lightning even captures it
- 0:44well enough really especially if you look back at the first one
- 0:47of these i did yes we were talking technology but we were not talking about
- 0:51the ai the way you know the way we're probably going to talk about it today
- 0:54right the pace at which it's moving is like nothing i've ever seen certainly
- 0:58right in my 30 years now in the industry wow so so so let's lay the field let like
- 1:05let's lay out the field. What has really happened?
- 1:09Like, what is happening now that you didn't really foresee? Or is this kind
- 1:12of like, hey, this is kind of where we thought it was going to go over these past four years?
- 1:16You think you see ChatGPT coming onto the scene in two, a little over two years ago now.
- 1:22Right, right. And that's really the sort of the seminal point that changed everything.
- 1:26Yeah, AI was around at the time, but it was more traditional machine learning,
- 1:30you know, that type of AI.
- 1:32And while it was improving and getting enhanced over time, it was not moving
- 1:37at the pace that we've seen generative AI go.
- 1:40Since, you know, like I said, ChatGPT came on the scene just over two and a half years ago.
- 1:45That's really been the catalyst that has sped things up to the level that we're
- 1:49seeing today. Huge push.
- 1:51So I know one of the concerns that firms have, and I think a lot of people have with.
- 1:56AI as data integrity and protecting your information, making sure that it's
- 2:00not used on other models.
- 2:02Can you talk about what's happening in that system and maybe what Dell Tech
- 2:06is doing or however you want to talk through what that looks like right now? That's a big concern.
- 2:11Right. And there's so many facets of that when you talk about data integrity,
- 2:14because part of it, as you said, making sure that proprietary data isn't being
- 2:20used to train getting outside of sort of the domain that you want it in.
- 2:24And the way I typically like to answer this question is we're in an era of hyper-focus
- 2:29on terms and conditions more so than ever before.
- 2:32Not that we should have ignored them in the past. Right.
- 2:34Boy, I think it's more important than ever to really understand how all these
- 2:38tools that are out there are leveraging data because, you know,
- 2:42some nuance in those terms and conditions,
- 2:44whether it's paid for version or a not paid for version, can have a big impact
- 2:48on how they're using your data and those types of things.
- 2:50So again, what I would say, hyper-focus on sort of terms and conditions more than ever before.
- 2:55Wow. Okay, so you wouldn't want your AI to say, hey, give me a quick summary
- 2:58of the terms and conditions because you may miss something, right?
- 3:03May miss something. That's great. I mean, talk about testing integrity, right?
- 3:06How smart is the AI to take it to its advantage? I don't know.
- 3:11Well, I think it's a good segue because then you think about,
- 3:13okay, we have this great AI model, this great AI system.
- 3:16Where do the humans come in, right? Where do they kind of, what are their touch points?
- 3:19What does that look like? What is, I hate to say it, the relationship between
- 3:23AI and a human and how that functions?
- 3:28Well, I think probably the best way to describe it is a human in the loop, so to speak.
- 3:33From my sort of experience in many different facets, the most successful application
- 3:38of AI is where there is a human in the loop.
- 3:41And what that means is, yes, AI can automate things. Yes, AI can serve up insights.
- 3:48Yes, AI can augment content creation. But obviously, to ever blindly take that
- 3:55as truth and fact and whatnot, you need to have a human in a loop to sort of
- 3:59press that last button to make that approval, whatever the case may be. Right.
- 4:02You know, all the things that that Deltec is doing in terms of automation and
- 4:06whatnot, we're making sure that there is a human in the loop,
- 4:09not only to make that that last decision, but also a way to review how decisions
- 4:14and insights were sort of established.
- 4:16Right. So if I need to say, well, this seems a little odd or off,
- 4:20I can look at, OK, how did the AI, if you will, come to this conclusion?
- 4:24And I can hopefully make it better the next time around.
- 4:27So, again, that that human in the loop thing, I think, is absolutely critical. Right. And it
- 4:32So if we think about human in the loop, and then I keep hearing this term agentic, right?
- 4:36The agentic future, the future is agentic or agent based.
- 4:42Is that what's happening? Is there going to be a screen with the agent and then
- 4:45you're the human kind of working and saying, do this, do this.
- 4:48Help me with this project. I mean, is there something, is there a use case or
- 4:51anything you want to talk about?
- 4:52Definitely. And it's interesting. You know, you think about this industry,
- 4:56AEC, architecture, engineering, construction, the built environment.
- 4:59You often hear the term digital twin. Yes. And in the context of the AEC industry, we know what that is.
- 5:05It's a twin, a digital twin of a built environment that you can use to monitor
- 5:09and test and explore and whatnot.
- 5:11Well, think of a digital twin in sort of your day-to-day life.
- 5:15You know, you've got this agentic AI being, if you will, that can do things
- 5:19for you, that can flag things for you, that can serve up information for you.
- 5:23That's one of the ways we're looking at it. You have your real self,
- 5:26but you also have your digital self that's doing work in the background for
- 5:30you and, you know, finding things that you may need to look at,
- 5:33course correct decisions on a project, you know, things that are popping up
- 5:37that it can tell you about or take action on.
- 5:40Gotcha. So it's right there to help you along your way, your digital twin.
- 5:45It's like your better half, hopefully not, right? It does its best to kind of replicate you, right?
- 5:52Um so i i think i'm ahead of your your education
- 5:55session today which is transforming project management how ai can
- 5:58streamline processes processes and boost productivity i
- 6:02think we covered the boost productivity part right you have a digital twin doing
- 6:04everything in terms of streamlining processes what does that look like well
- 6:09it can come in so many forms and that's sort of a a key theme of of the session
- 6:14that i'm delivering in that you know you look at And it can be intimidating.
- 6:17It's this magical thing that's out there.
- 6:20But the reality is, it's one of the most accessible technologies that I've sort
- 6:25of come across in my business life, if you will.
- 6:28And I think that's important for people to understand because when you talk
- 6:31about increasing productivity,
- 6:33when you talk about streamlining processes and those types of things,
- 6:36the tools that are available, you and I can walk out of this podcast today and go sign up,
- 6:41use one of those tools, whatever the case may be, and be more streamlined within
- 6:45a matter of time. of hours, if not minutes in many, many cases.
- 6:48And I think it's important for people to understand and look at it that way
- 6:52because you can easily get scared off, if you will, or intimidated because,
- 6:57you know, this magical, mystical AI thing is out there and I don't really understand how it can help me.
- 7:03You don't need to know everything about AI. All you need to do is really get
- 7:07a start with it and you'll start to see results.
- 7:09So you're saying like, hey, get started, give it a shot. What's,
- 7:13if I'm an engineering firm and I say, hey, I'm ready to start investing in AI.
- 7:17I want to take the leap. I've heard all the, I've read the case studies.
- 7:21I've done my research. I'm ready to get started. What is the,
- 7:24should I just dip a toe in?
- 7:26Kind of take it slow. Obviously it sounds like we don't want to dive in,
- 7:28but you want to understand the,
- 7:30where you're going and almost have a roadmap for your own firm.
- 7:34Yeah, yeah. I think there's probably two things to look at in that regard.
- 7:37First is look at the tools you already have in place.
- 7:40Chances are you probably have technology in place that has AI embedded in it,
- 7:44whether it's Microsoft Copilot, whether it's some of the stuff that we've embedded
- 7:48in our Dell Tech solutions.
- 7:50You know, start with those types of tools because they really work well to demystify
- 7:55it for your staff in terms of, hey, I'm already using AI and I'm using it in
- 7:58a pretty cool, powerful, helpful way. So I think that would be my first point.
- 8:02My second point is, you know, find those small use cases, find those problematic
- 8:07areas that are impacting the business operationally, whatever the case may be.
- 8:12Typically, I suggest starting with an internal focused one, not a client focused one. Right.
- 8:16But find those use cases, those simple use cases that you think you might be
- 8:19able to solve with AI, find the tool, apply it to it, and then use that as sort
- 8:24of foundational wind to build on from there.
- 8:27I think those are sort of the two things that if you want to,
- 8:29you know, dip your toe in and get started, that's that's the great sort of a
- 8:32great place to start. Okay.
- 8:34So let's say that you're already in AI, right? You're, you're kind of,
- 8:37you're further down the path than just dipping your tail.
- 8:40And you're like, man, where do I, where do I go from here? I feel like I've
- 8:42done everything I can with the tool, or I feel like I'm, I'm at a roadblock.
- 8:47Is that something that you're seeing? Is that like, are there other ways to
- 8:51go with it? Or is it all encompassing? Is there, is there even an answer to that question?
- 8:55I think, I think there's lots of answers probably. And it just kind of depends
- 8:59on the company and whatnot.
- 9:01I, what I would say is there's, I think, two things there, too. First of all, data.
- 9:04You know, you've heard the cliche, data's the new oil.
- 9:07That's even more the case in the AI-driven world because, let's face it,
- 9:12the AI tools that we're going to use tomorrow are going to be driven by the
- 9:16data that we're creating today.
- 9:17So one thing you need to think about is, how are you managing data within your
- 9:22business? And this can come in many different forms.
- 9:24One, use the tools that you have in place the way they were intended to be used,
- 9:28because if you do that, chances are you're creating data that's going to enable
- 9:31some AI capability down the road for yourself.
- 9:34But then I think the other thing is start to do an analysis of what data is important to your firm.
- 9:40And that can be anything from sort of knowledge that you want to capture from
- 9:43the people within your business,
- 9:44your own experience from past projects that you delivered, your domain expertise,
- 9:49and start to figure out how you can capture that in a way that you can serve
- 9:53it up through some sort of AI platform, which, again, are becoming...
- 9:57Capability is becoming more and more accessible every day. Yeah,
- 10:01I think accessibility is a key word here because it's easy to get started, like you said.
- 10:06So, Brett, thanks again for joining the show. I want to give you the last word
- 10:08here. Is there anything else that our viewers or listeners should know about
- 10:12what's happening in the, we'll call it the environment of AI?
- 10:16So we've talked about human in the loop.
- 10:18I think it's important to talk about those humans because we're at a time,
- 10:23you know, there's all this rhetoric around, you know, AI is going to replace
- 10:25people and eliminate jobs and all that kind of stuff.
- 10:28And of course, the cliche follow-up to that is AI is not going to replace humans.
- 10:33AI, humans that know AI are going to replace humans that don't.
- 10:37And I think every company needs to be aware of that.
- 10:40There's a, there's sort of the need for a hybrid worker, specifically in the
- 10:44engineering industry that not only has the engineering domain expertise,
- 10:48but that digital fluency so that they can be that human in the loop working with AI,
- 10:54but still have that industry domain expertise to bring to the table as well.
- 10:59So I think, you know, how we look at humans in terms of the knowledge workers
- 11:02coming into our business is going to need to change a little bit because we're
- 11:06looking for different qualifications than we used to be.
- 11:08That's an interesting point. I know we're going to wrap it up,
- 11:10but let's talk workforce just for a moment, because I think that plays into
- 11:13what's happening from a workforce perspective within the industry.
- 11:18Trying to fill new engineering jobs, backfill has
- 11:21been it's it's it's an issue right
- 11:24now so is ai a partial
- 11:28solution to that problem in that hybrid worker that you're you're talking about
- 11:32as you said i think you said it right it's a partial solution i don't think
- 11:35it's the solution but it's definitely a partial solution okay there's no doubt
- 11:39ai can automate things to um require a few less resources that doesn't eliminate
- 11:45the need for domain expertise,
- 11:47but definitely can streamline things to, you know, maybe lighten the load for
- 11:53those positions that we're having
- 11:54trouble filling because of the workforce challenges that are out there.
- 11:57Right, right. Well, Brett, again, thank you so much for joining the show for a fourth time.
- 12:02I hope to see you at Fall Conference in San Diego. You can't turn on an invite
- 12:05to that. We'll talk again.
- 12:07Thank you very much for having me. Thank you, Brett. And again,
- 12:09this has been an episode of Engineering Influence, a podcast from the American
- 12:13Council of Engineering Companies. Make sure to like, subscribe,
- 12:16and follow us on all of our channels.
- 12:17Music.