How NVIDIA Sees AI Transforming AEC

WEBVTT
00:00:01.328 --> 00:00:06.152
A big thank you to Autodesk, the sponsor of Engineering Influence at this year's convention.
00:00:06.692 --> 00:00:11.911
Autodesk partners with the AEC industry and its leaders to shape a more connected,
00:00:12.364 --> 00:00:17.487
data-driven future, empowering engineers to work more intelligently and to make
00:00:17.487 --> 00:00:20.137
confident decisions to drive meaningful impact.
00:00:20.543 --> 00:00:24.798
This enables firms to evolve, stay competitive, and deliver stronger business
00:00:25.152 --> 00:00:28.020
outcomes, all while designing a better world.
00:00:28.293 --> 00:00:30.241
So thank you again to Autodesk.
00:00:31.160 --> 00:00:34.465
Welcome to the Engineering Influence Podcast from the American Council of Engineering
00:00:34.465 --> 00:00:38.724
Companies, coming to you from our 2026th Annual Convention and Legislative Summit.
00:00:39.130 --> 00:00:43.298
One of the issues that has been really burning for a couple of years now,
00:00:43.704 --> 00:00:45.521
we were at really the outset of it.
00:00:45.939 --> 00:00:50.670
Actually, I take that back because I was at FIDIC in Mexico City in 2019,
00:00:51.135 --> 00:00:55.785
and we had a former Silicon Valley exec who said that if you're at an engineering
00:00:55.785 --> 00:00:58.925
firm or you're at a data firm, you just don't know it yet, that was 2019.
00:00:58.925 --> 00:01:01.665
And now I could say that if you're an engineering firm, you're actually at an
00:01:01.665 --> 00:01:03.250
AI firm, you just don't know it yet.
00:01:03.848 --> 00:01:10.865
I'm joined by Sean Young, who runs really the AEC component of NVIDIA with the
00:01:10.865 --> 00:01:14.805
Institute on its Firm of the Future initiative. Sean, thank you for joining
00:01:14.805 --> 00:01:15.896
the program. Yeah, thanks, Jeff.
00:01:16.337 --> 00:01:19.785
I mean, you know, it was funny back then. They said, well, you know,
00:01:19.785 --> 00:01:24.405
if you're at a data firm, you just don't realize it. And everyone's like, oh, OK.
00:01:24.852 --> 00:01:31.265
And from 2019 to about 2020, now, 2026, now it's shifted completely to now it's
00:01:31.265 --> 00:01:34.775
like you are now at an AI firm. You just don't know it yet.
00:01:35.666 --> 00:01:40.145
How, it's not really a question of if, but when. From your perspective,
00:01:40.145 --> 00:01:46.493
you know, how is AI really changing the landscape in the AEC industry?
00:01:46.800 --> 00:01:51.175
You know, it's, what's interesting as you're talking, I'm thinking what you
00:01:51.175 --> 00:01:53.158
say can be applied to every company. Yeah.
00:01:53.561 --> 00:01:57.365
Cause like, it's like, how did you operate before you had a computer?
00:01:57.365 --> 00:02:00.829
How did you possibly, what you use a fax machine and telephone?
00:02:01.449 --> 00:02:06.586
How, we're going to say that about AI. How did you operate before AI?
00:02:06.981 --> 00:02:11.350
And AI is not, in the near term, something that's going to be spoon-fed.
00:02:12.310 --> 00:02:20.700
It's not as simple as straightforward as the evolution from paper to computers with CAD was.
00:02:20.700 --> 00:02:24.940
Because that was just like, go buy a computer, the software company will tell
00:02:24.940 --> 00:02:28.520
you what computer to buy, and you double-click on the installer and installs
00:02:28.520 --> 00:02:31.160
your thing, and you just read our manual. well, here's the things you can do
00:02:31.170 --> 00:02:32.218
and here's the things you can't do. Yeah.
00:02:32.438 --> 00:02:36.190
AI can do just about anything. Yeah. Which is where it gets complicated because
00:02:36.570 --> 00:02:44.025
it means you have to be the investigator and the arbitrator of that knowledge within your firm. Yeah.
00:02:44.158 --> 00:02:49.390
And there's a lot of knowledge out there. The good thing is we can all use large
00:02:49.390 --> 00:02:54.480
language models, chatbots to help us navigate all that knowledge so you can
00:02:54.480 --> 00:02:58.270
start using AI to help you figure out how to use AI. Yeah.
00:02:58.970 --> 00:03:03.536
Yeah, you can kind of like help it build the ship while you're trying to sail at the same time.
00:03:03.716 --> 00:03:07.520
And everybody kind of focuses on the large language models that are out there
00:03:07.520 --> 00:03:10.914
and pervasive, of course, in everything that we're seeing now.
00:03:11.332 --> 00:03:15.670
But it goes beyond that. I mean, it really is a question of how do you take
00:03:15.670 --> 00:03:19.630
all this information that you own as a company? How do you leverage that information
00:03:20.007 --> 00:03:23.972
into an AI model that actually produces the desired outcome?
00:03:24.448 --> 00:03:28.030
One thing I was really interested in talking about this issue,
00:03:28.490 --> 00:03:32.480
you know, a couple of years ago were how few firms in the AEC space actually
00:03:32.480 --> 00:03:36.525
had people in the C-suites who were actually focused on data.
00:03:36.816 --> 00:03:40.639
Either a chief data security officer, chief information officer.
00:03:40.987 --> 00:03:44.160
I mean, it goes, you know, everybody had an IT department, but nobody was actually
00:03:44.160 --> 00:03:47.480
thinking holistically or universally about how their data is maintained and
00:03:47.480 --> 00:03:49.930
security of it. And then how do you actually take advantage of it?
00:03:50.868 --> 00:03:54.903
Yeah. What have you seen in the industry over the past couple of years about
00:03:54.903 --> 00:03:58.453
how that's changed? In this industry, not a whole lot.
00:03:58.933 --> 00:04:01.273
In other industries, there's data scientists.
00:04:01.812 --> 00:04:04.943
And so firms have hired data scientists to organize all the data.
00:04:04.943 --> 00:04:10.717
Because in this industry, very services based, the project's done, you archive the data.
00:04:11.600 --> 00:04:16.693
In other industries, you learn from that data. So the data, like in healthcare,
00:04:17.293 --> 00:04:20.125
the data is informing all sorts of research and discovery.
00:04:20.508 --> 00:04:22.453
Well, it could be the same with AEC.
00:04:23.073 --> 00:04:28.693
You just need to harness that. So what I would say is now, you know, your data is your IP.
00:04:29.204 --> 00:04:33.733
It's the way you do things. The data is in all sorts of different files.
00:04:33.733 --> 00:04:38.655
Some of those files are software vendors files and some of it's Excel or PDFs or whatever.
00:04:39.137 --> 00:04:43.173
But that's okay. It used to be that you would have to get, I mean,
00:04:43.173 --> 00:04:45.833
used to be like the ancient history about a year ago. Yeah, exactly.
00:04:45.833 --> 00:04:49.093
I have to go get a data team of data scientists to figure out how to crack all
00:04:49.093 --> 00:04:54.096
the different data formats, normalize all the data into one homogenous data lake, blah, blah, blah.
00:04:54.386 --> 00:04:57.422
But now you can just have agents go and figure it out.
00:04:57.973 --> 00:05:02.438
Just tell an agent, you go, you know, get, here's all of my data at NVIDIA.
00:05:02.954 --> 00:05:07.903
We have tons of data repositories, but we have one, well, actually I have many
00:05:07.903 --> 00:05:11.545
agents, but all of those agents have access to all the data.
00:05:11.765 --> 00:05:15.163
And it just goes, it doesn't matter, I don't know what format it's in.
00:05:15.563 --> 00:05:19.689
I just give it a prompt and access to all of the repositories.
00:05:20.026 --> 00:05:24.223
It goes and fetches all the data. Well, I found this and this in Confluence
00:05:24.223 --> 00:05:29.151
and I found this in GitHub and I found this and here's what you got.
00:05:29.843 --> 00:05:32.313
And that, I think, is a good transition from looking at, like,
00:05:32.313 --> 00:05:35.883
okay, everybody knows about the large language models, and you kind of prompt
00:05:35.883 --> 00:05:37.342
something, you'll get a result.
00:05:37.719 --> 00:05:42.843
If you have a closed data set, you can train it and create a project that actually
00:05:42.843 --> 00:05:48.213
has those parameters to say that you must search within this data set with these limitations.
00:05:48.573 --> 00:05:50.343
That's one part of the process.
00:05:50.726 --> 00:05:53.923
What we're looking at in the future is really the spread of agentic AI.
00:05:53.923 --> 00:05:57.563
And let's boil that back for people who aren't focusing on it as much as you
00:05:57.563 --> 00:05:59.643
and I are. what's an agent,
00:06:00.307 --> 00:06:06.367
And how can it be employed by an ADC firm to actually do work faster?
00:06:06.861 --> 00:06:11.532
Okay, well, everybody, I'm assuming, is familiar with ChatGPT at this point.
00:06:11.817 --> 00:06:15.317
So ChatGPT is a form of a large language model.
00:06:15.608 --> 00:06:20.333
It's a model that was trained on very generic data, all of the internet,
00:06:20.617 --> 00:06:22.307
and it understands languages.
00:06:22.907 --> 00:06:26.713
One of those languages could be coding. One of those languages could be English, et cetera.
00:06:27.003 --> 00:06:30.597
In that language understanding, there's also an understanding of all the context
00:06:30.597 --> 00:06:33.406
of history and different subject matters.
00:06:34.147 --> 00:06:40.363
So, but the method of you interacting with ChatGPT is you type something in, which is a prompt.
00:06:40.770 --> 00:06:45.177
It processes that and gives you a response, which is, by the way,
00:06:45.177 --> 00:06:46.877
what it's doing is inference. Yeah.
00:06:46.999 --> 00:06:50.760
And it generates tokens. Just side note, generates tokens.
00:06:51.027 --> 00:06:52.612
And tokens are really the.
00:06:54.907 --> 00:07:02.167
You know, the currency of AI. A token is essentially your action within that model.
00:07:02.399 --> 00:07:10.960
The token is the unit of compute that the inference model is generating to give you a response.
00:07:11.547 --> 00:07:16.607
So your response may be one token, depending if the inference requires some
00:07:16.607 --> 00:07:19.656
reasoning or some thinking, it may be multiple tokens.
00:07:20.028 --> 00:07:24.648
And each token has a cost. And as things develop and things get better,
00:07:25.072 --> 00:07:29.937
the price per compute is going to essentially lower, and then you'll have,
00:07:30.088 --> 00:07:34.062
yeah, greater flexibility in time and theory.
00:07:34.166 --> 00:07:36.927
Yeah, it also depends where you get your computer. Exactly. Because if you're
00:07:36.927 --> 00:07:41.347
running your models locally on hardware that you own, then you have infinite
00:07:41.347 --> 00:07:45.701
amount of tokens until, I mean, you run out of compute space. Yes.
00:07:46.130 --> 00:07:52.807
Okay, so back to what is egetic. So in ChatGPT, you type in a prompt and you get a response.
00:07:53.056 --> 00:07:58.484
Then you learn something and you type in a prompt and you get a response. Well, what if you're...
00:07:59.047 --> 00:08:04.327
Question was how do i use my cad program to to do this thing.
00:08:05.267 --> 00:08:09.797
And the response would be here are the 30 steps or here are the 100 steps or
00:08:09.797 --> 00:08:13.207
here are the 1000 steps yeah and you'd be like all right i'll copy and paste
00:08:13.207 --> 00:08:14.727
that into a document and then go do those,
00:08:15.367 --> 00:08:20.217
1000 1000 things now i could do this cool thing i never knew how to do before
00:08:20.217 --> 00:08:23.257
and it's customized to specifically what i wanted to do and i didn't have to
00:08:23.257 --> 00:08:27.917
read the manual pretty awesome except you still have to do the 1000 steps exactly
00:08:27.917 --> 00:08:29.272
So you're not that far ahead.
00:08:29.736 --> 00:08:33.637
Well, where Agentic comes in is now you say, all right, do that for me.
00:08:34.839 --> 00:08:42.351
And now the computer, the Agentic AI is going into that application and doing it all for you. Yeah.
00:08:42.705 --> 00:08:45.227
And you just skipped a thousand, and you can go to sleep. Exactly.
00:08:45.567 --> 00:08:48.487
And that's why I think one of the big takeaways in the Firm the Future Research
00:08:48.487 --> 00:08:53.967
and the report that is just now out that essentially AI is not a question of replacing engineers.
00:08:53.967 --> 00:08:57.525
It's going to be allowing engineers to do things faster, more efficiently,
00:08:57.961 --> 00:09:04.084
and to free them up to do the things that they really should be doing by leveraging agents to,
00:09:04.674 --> 00:09:09.324
do all these sub-operations at a level.
00:09:10.139 --> 00:09:13.247
A multiplication effect that was impossible beforehand. So you could have a
00:09:13.247 --> 00:09:17.077
firm of 20 engineers that now becomes unlimited by the number of agents that
00:09:17.077 --> 00:09:20.785
you're actually allowed or can produce to do the work. That's right.
00:09:20.913 --> 00:09:23.078
Every engineer now has a team.
00:09:23.953 --> 00:09:28.843
Resources are unlimited as long as you have compute. So you have a team of 10 people.
00:09:29.507 --> 00:09:32.647
Before I start my day now, before I go in and check my email,
00:09:32.647 --> 00:09:36.192
I want to get my agents rolling. I want to get at least a couple agents rolling.
00:09:36.488 --> 00:09:41.217
So like they're doing work while I'm doing work. because once I get into email
00:09:41.217 --> 00:09:43.036
and meetings, I mean, I have to get them rolling.
00:09:43.309 --> 00:09:47.847
But I have some, I just had a team call. Some people on my team have automatic
00:09:47.847 --> 00:09:49.867
agents, so they're just working all the time.
00:09:50.327 --> 00:09:52.580
They don't have to get them started.
00:09:53.467 --> 00:09:58.947
So, and those agents could be, I mean, they could be doing a lot of work for
00:09:58.947 --> 00:10:02.637
you, but also helping you iterate because there's experimentation involved in
00:10:02.637 --> 00:10:04.029
the process. There's creativity.
00:10:04.308 --> 00:10:08.697
Well, I'll have this agent try 10 things this way. I'll have another agent try
00:10:08.697 --> 00:10:11.667
another 10 things the other way, and let's see which one wins.
00:10:12.067 --> 00:10:16.827
It allows firms that don't have this. It used to be that you'd have to be a
00:10:16.827 --> 00:10:20.597
very large firm with a lot of different capabilities to be able to do the amount
00:10:20.597 --> 00:10:21.767
of iteration, the amount of,
00:10:22.287 --> 00:10:25.766
experimentation that, you know, there's no way a small firm could do it.
00:10:25.952 --> 00:10:28.936
But leveraging agentic AI...
00:10:29.630 --> 00:10:34.745
Setting it up properly, it really does completely level the playing field that
00:10:34.745 --> 00:10:41.356
you can now engage in the same amount of testing iteration that any of the large ones get. That's right.
00:10:41.577 --> 00:10:46.835
And like, you know, in terms of replacing people, I mean, that's not high value
00:10:46.835 --> 00:10:49.466
work, in my opinion, pushing buttons in a software application.
00:10:49.773 --> 00:10:56.296
Like the high value is bringing your ingenuity to bear and applying your creativity and making decisions.
00:10:56.575 --> 00:11:01.863
But like have the agents go and do the boring stuff that is tear-pulling.
00:11:02.955 --> 00:11:06.025
Yeah, it's empowering the engineer to do the work that they were trained to
00:11:06.025 --> 00:11:07.118
do. Yeah, that they want to do.
00:11:07.327 --> 00:11:10.795
Now, looking ahead, I mean, from NVIDIA's perspective, I mean,
00:11:11.095 --> 00:11:16.985
both from just the development of AI, both in hardware and then the implementation,
00:11:16.985 --> 00:11:22.073
I mean, where do you, I mean, you're engaged with the firm of the future.
00:11:22.677 --> 00:11:26.461
It's fantastic to have a company like NVIDIA part of this process to help guide the future.
00:11:27.835 --> 00:11:29.117
What is, you know, NVIDIA's.
00:11:29.903 --> 00:11:34.894
Entrance into the EEC space? I mean, where do you see the growth opportunity there?
00:11:35.214 --> 00:11:40.024
Where is it from your company's perspective on what the future holds in our
00:11:40.024 --> 00:11:44.302
industry? Ah, it's a really open-ended question. I know.
00:11:44.505 --> 00:11:48.594
The sky's the limit. The sky's the limit. And where NVIDIA fits is we're enabling
00:11:48.914 --> 00:11:53.420
the ecosystem to do better, to do more with less, to accelerate.
00:11:53.699 --> 00:11:58.214
We work with all the software companies to help them at the low level with these
00:11:58.214 --> 00:12:05.022
computationally complex processes that can be accelerated on our technology stack.
00:12:05.394 --> 00:12:08.554
So that's AI is a great example.
00:12:08.554 --> 00:12:14.504
We have, yes, all these applications hit our hardware, but we have tons of software
00:12:14.504 --> 00:12:19.454
libraries for developers that they implement and integrate into their software
00:12:19.454 --> 00:12:21.856
and use to develop their AI,
00:12:22.239 --> 00:12:24.886
and that gets rolled out on our interviews.
00:12:25.297 --> 00:12:27.864
A portion of the firm of the future also looks at, of course,
00:12:27.864 --> 00:12:31.734
the future of the firm in 2035. And, of course, we look at the workforce issue
00:12:31.734 --> 00:12:37.194
and really what is the firm going to be after we get to the point where AI,
00:12:37.194 --> 00:12:39.118
energetic AI becomes kind of ubiquitous.
00:12:39.466 --> 00:12:43.224
And you talk about pretty much the AI, you know, the AI first mentality.
00:12:43.224 --> 00:12:47.124
It's not a firm that implements AI in the back end, but it's one that actually
00:12:47.124 --> 00:12:51.294
is built on the foundational model of having AI at its core.
00:12:51.814 --> 00:12:58.156
Right. You know, how do you see the industry evolving to be an AI first industry?
00:12:58.790 --> 00:13:05.413
I think the industry needs to first think about itself as a technology-first industry.
00:13:05.882 --> 00:13:10.370
Because I think it's been like a technology laggard. Yeah.
00:13:10.903 --> 00:13:18.053
Not taking, I would say, firms in general don't really take accountability for
00:13:18.053 --> 00:13:20.348
technology. They buy off the shelf. Yeah.
00:13:20.673 --> 00:13:24.613
Which is fine. You could buy off the shelf, and you should continue to buy off
00:13:24.613 --> 00:13:26.510
the shelf. There's tons of options out there.
00:13:26.783 --> 00:13:30.806
But I think you really have to know what those options are and have to have.
00:13:31.368 --> 00:13:38.833
Particularly now with AI, have to have some ability to use your data,
00:13:38.833 --> 00:13:42.426
to leverage your data, to inform what those applications are doing.
00:13:42.675 --> 00:13:47.503
So, for example, if you buy a software application that has AI functionality
00:13:47.503 --> 00:13:51.653
in it, it's generic. Just the same way Chad GPT is generic. It's trained on
00:13:51.653 --> 00:13:52.758
the entirety of the internet.
00:13:53.031 --> 00:13:58.963
So some AEC-specific AI feature that is from a software company is trained on
00:13:58.963 --> 00:14:00.792
some data set that wasn't yours.
00:14:01.018 --> 00:14:03.323
It doesn't reflect the way your business works.
00:14:03.613 --> 00:14:09.743
So the question becomes, how do you then teach that AI to do things Jeff's way?
00:14:10.183 --> 00:14:14.381
Because Jeff's been doing, Jeff Incorporated has been doing the same thing,
00:14:14.590 --> 00:14:19.163
building a hospital in this region according to this building code and the way
00:14:19.163 --> 00:14:23.583
this region or hospital board likes to work for 50 years.
00:14:23.943 --> 00:14:27.003
And you know what you're doing better than anybody. That is not going to be
00:14:27.003 --> 00:14:30.785
encapsulated in that generic AI. So you need to teach that AI with your data.
00:14:31.023 --> 00:14:34.423
In order to do that, you have to have people in your firm to understand all this technology.
00:14:34.843 --> 00:14:38.103
And that's the biggest thing I think is, you know, even organizationally from
00:14:38.103 --> 00:14:44.534
AECC, We recognize this and we realize that, you know, we have a lot of data.
00:14:45.421 --> 00:14:48.452
It's not structured appropriately to be able to leverage AI.
00:14:48.452 --> 00:14:53.022
You have to get to the point where you actually understand the entire universe
00:14:53.022 --> 00:14:57.182
of your data and then structure it in such a way where you could put it into
00:14:57.182 --> 00:14:59.916
a model to actually have it do what you want to do.
00:15:00.912 --> 00:15:04.067
So that's really interesting because I would have totally agreed with you a year ago.
00:15:05.152 --> 00:15:08.772
Is that all now outmoded thinking or is it? I think we're getting there.
00:15:08.772 --> 00:15:11.968
I mean, there's two schools of thought. One is exactly what you just said.
00:15:12.258 --> 00:15:17.532
You got to go in brute force, get some data scientists, get all of your data
00:15:17.532 --> 00:15:23.955
in one repository, normalize it all into the same format, homogenize it all.
00:15:24.164 --> 00:15:27.992
And then all of a sudden you have a single point of view. And if you think about
00:15:27.992 --> 00:15:30.363
tabular data, you can zero in on exactly what you need.
00:15:31.002 --> 00:15:34.253
But with agentic AI- You can have it do it for you. Yeah.
00:15:34.828 --> 00:15:38.501
Just put it in and say, okay, here's the data, sort it by kind,
00:15:38.501 --> 00:15:41.798
sort it by X, Y, Z, whatever kind of tag you want to put it in.
00:15:41.914 --> 00:15:44.761
You figure it out. Figure it out and then come out. Yeah, you figure it out. So there you go.
00:15:44.761 --> 00:15:48.721
So I'm behind the times already because I'm trying to figure out like, you know, everybody is.
00:15:48.721 --> 00:15:52.089
And that's the thing. This is happening so fast. But the point is,
00:15:52.385 --> 00:15:55.938
you know, all of a sudden you have to worry a lot less about interoperability
00:15:56.060 --> 00:15:58.571
if you're not having to brute force a situation.
00:15:58.571 --> 00:16:01.673
Absolutely. If you have agents that are working around the clock to just go,
00:16:01.940 --> 00:16:06.361
you know, whether it's import, export, normalize, find some standards and normalize
00:16:06.361 --> 00:16:12.134
around that, whatever it is, agents are going to figure it out and the agents can figure it out.
00:16:12.771 --> 00:16:15.599
And you just keep testing until you find like a non-lossy way to do it.
00:16:15.727 --> 00:16:18.471
So we could talk about this for hours.
00:16:18.831 --> 00:16:22.176
I think that, you know, the Firm the Future report that came out,
00:16:22.554 --> 00:16:25.451
you know, this week, I encourage everyone to take a look at it.
00:16:25.451 --> 00:16:27.824
It's up on the ACC Research Institute website.
00:16:28.260 --> 00:16:32.485
It's the continual research product coming out of this initiative.
00:16:32.972 --> 00:16:37.131
What's next for the work that the Institute's doing and NVIDIA's work with the
00:16:37.131 --> 00:16:41.523
Institute on kind of the next product in that firm of the future?
00:16:41.819 --> 00:16:46.291
Well, what I'm super excited about is, first of all, all this stuff can be done
00:16:46.291 --> 00:16:48.891
today. And so we talked about ChatGPT.
00:16:49.621 --> 00:16:55.051
That's OpenAI. Yeah. And OpenAI has an agentic toolkit called Codex.
00:16:55.051 --> 00:17:00.680
Yep. And you can go and download it and start using it. Got it on my Mac. Right on. And so now...
00:17:01.777 --> 00:17:06.247
Back to that automation system instead of giving me the 1000 lines of code just
00:17:06.247 --> 00:17:12.017
tell your codex to go and do it the way it does it is by into with a traditional
00:17:12.017 --> 00:17:15.337
isv application is by connecting to that application
00:17:15.717 --> 00:17:20.817
through the applications api yeah all applications have apis which is like if you're a programmer,
00:17:21.277 --> 00:17:25.407
or a plugin developer you can write some code that accesses the application
00:17:25.407 --> 00:17:27.778
through the api yeah well now your codex
00:17:27.970 --> 00:17:31.547
can access the application so once you have that It's connected,
00:17:31.547 --> 00:17:34.687
and then you can just create the agent to do exactly what you want through that
00:17:34.687 --> 00:17:36.457
API connection to that app.
00:17:36.978 --> 00:17:40.798
And essentially, that's a blank canvas for whatever you need to accomplish. That's right.
00:17:41.024 --> 00:17:46.637
And now, to make it even easier, so the other example of an agentic tool is
00:17:46.937 --> 00:17:49.851
Claude Cowork. And Cowork just...
00:17:50.453 --> 00:17:54.174
Last week, announced a bunch of connectors for creative workflows.
00:17:54.865 --> 00:17:59.595
SketchUp, Autodesk Fusion, Blender. So it's all coming together.
00:17:59.595 --> 00:18:03.605
It's all starting to come together. So that's the MCP and the skills that it
00:18:03.605 --> 00:18:06.991
needs, all that to go and do the work for me.
00:18:07.229 --> 00:18:10.718
Now, I mentioned three applications. Why couldn't you just tell,
00:18:11.083 --> 00:18:14.725
you know, co-work, hey, go do something with all those three applications.
00:18:15.251 --> 00:18:18.647
So use all the applications you want. They're all installed on my computer.
00:18:18.937 --> 00:18:22.865
So on my team, we're testing now, here is the RFP.
00:18:23.985 --> 00:18:31.325
Generate an Excel version of the RFP with all the key data. Go and use Rhino with Grasshopper.
00:18:31.805 --> 00:18:35.032
Create a conceptual drawing of the building with some options.
00:18:35.300 --> 00:18:39.154
Maybe even I'll give it a sketch as some inspiration.
00:18:39.538 --> 00:18:42.418
Generate a conceptual model. Render that for me.
00:18:42.886 --> 00:18:45.382
Bring it into a BIM application.
00:18:45.939 --> 00:18:50.873
Create a structurally sound model based off of all that and create the construction docs
00:18:51.268 --> 00:18:55.215
one prompt and i was just on a call with my team and we were testing this out
00:18:55.215 --> 00:19:01.015
and watching it and we're like wait a second put 49 columns how did he get that
00:19:01.015 --> 00:19:04.925
information is that in the rfp like how does it know to do 49 columns and,
00:19:05.405 --> 00:19:07.835
and we went and checked the rfp that we didn't there's nothing like that in
00:19:07.835 --> 00:19:12.175
the rfp so somehow somehow just figure that out and that's funny and it shows
00:19:12.175 --> 00:19:15.815
how quickly things were developing because I've had conversations with firms
00:19:15.815 --> 00:19:20.009
that were on a leading edge of AI and they were even talking about the fact that.
00:19:20.706 --> 00:19:23.522
Yeah, we haven't got to drawing yet. We're getting close to that.
00:19:23.522 --> 00:19:27.943
But now we're at that point where we can put the prompt in and actually render.
00:19:28.383 --> 00:19:33.522
And that was not too long ago. I mean, if you don't get on the train now,
00:19:33.522 --> 00:19:36.732
you're going to be left at the station. Yeah. Here's your prompt.
00:19:36.998 --> 00:19:44.295
Read my RFP, generate conceptual model, BIM model, conceptual rendering,
00:19:44.937 --> 00:19:49.204
and full set of construction docs. Yeah. These specific sheets that you're looking for.
00:19:49.587 --> 00:19:53.587
Well, let's wrap it here because I know you've got another education session to get to.
00:19:53.889 --> 00:19:57.732
You were at one earlier. It was standing room only. You were on stage earlier
00:19:57.732 --> 00:20:01.912
talking about this with the Firm of the Future group. I encourage everyone out
00:20:01.912 --> 00:20:04.785
there, again, look at the research, get involved with it.
00:20:05.121 --> 00:20:08.983
I'd love to have you on again as this keeps on developing.
00:20:09.453 --> 00:20:12.635
But for right now, we'll leave it there, Sean Young, NVIDIA,
00:20:13.262 --> 00:20:16.202
part of the Firm of the Future initiative over here at the ACS Research Institute.
00:20:16.202 --> 00:20:19.224
Thanks for your time. Thanks, Jeff. Really appreciate it. Thank you. Thank you so much.
00:20:19.375 --> 00:20:22.882
And again, that has been Engineering Influence, a podcast from the American
00:20:22.882 --> 00:20:24.863
Council of Engineering Companies. And check out.