Episode Transcript
[00:00:00] Speaker A: If you're going to build a new factory, you've got a lot of things you need to think about. One, you need to select the equipment that you're going to put into this space. And even preparing the building for that stuff could take a long time. The benefit is that you get much more control over what you're building. But the timeline can be longer and it can end up costing you a bit more in the end where like if you were to go refurbish, you could buy a building that was already built for that use case. You may also benefit from any existing infrastructure or equipment that's there.
And usually you can shave months, if not years off of the timeline.
[00:00:43] Speaker B: Before.
[00:00:43] Speaker C: We get into it here, Greg, it was your birthday since our last episode. For those of you that don't know, they're listening at home, Greg is Canadian and because of that he has informed us about the ways of the Canadians. Canadian tuxedos. Greg, was there any Canadian themes to your birthday? Break it down for us really quick.
[00:01:00] Speaker B: Well, maybe I'll instead of telling, I'll just show. And so maybe this gives the Canadian theme a little flavor here. This is the the local Diagon team and they're in their ski themed outfits that we had during the party. So when I turned 40 we did a Canadian tuxedo party and then this one was a ski themed party in California in May.
[00:01:20] Speaker C: What I love about this Will, is that Greg is literally in balmy San Carlos right now and and the idea of wearing ski attire is probably gonna make Greg break a sweat mid podcast while we're wearing all of the ski things in that photo.
[00:01:35] Speaker B: Well, I'll let Will go, but my parents were also here and they thought it was hilarious that we bought a snow machine from Amazon and paid money to make snow in Canada.
That's not a thing to say the least.
[00:01:49] Speaker A: It was totally worth it. It's also worth mentioning Greg is the oldest of all of us, so I'm not sure how many people knew that, but I think it's worth mentioning.
[00:01:56] Speaker B: I don't think I realized that Will, but really appreciate you.
[00:02:00] Speaker C: Well, I think that in itself is a great transition. So everyone listening at home now that they've got a good age on everyone here. And also shout out to I believe Greg's dad because you were listening as well. Good evening, good morning, good afternoon. If you're listening. This is the Movers and Makers podcast where we dive into all things manufacturing news, equipment and procurement news. With us today is the CEO and co founder of Diagon Will Drury and also the Vice president of business development and other co founder Greg Smythe and also resident Canadian tuxedo wear.
Gentlemen, we have a bunch of great topics to dive into. Hopefully none of them involve pulling up that photo of us wearing ski attire in San Carlos again. But we're going to talk about everything from vibe coding in the world of equipment and manufacturing and industrial. We're going to talk about reshoring. Greg dropped a little bit of a nugget last time. Reshoring needs a rebrand. So we're going to dive into that a little bit. And then I think the one thing that I think is really exciting is there been so much news around new factories just in the last week. So we're going to break it down a bit, some of that news and then talk about what it looks like to build some of those factories. So first off, this idea of Vibe coding. So Vibe coding is reaching a true moment of hype, I think across all software as a service businesses.
Vibe coding is kind of this art and science of you're using AI as a copilot, as a thought partner, as a sounding board as you're coding. But I've got to hear both of your takes.
What do you think about Vibe coding as it relates to the person in 2010 who was sitting behind a 5 axis Fanuc fiddling with a PLC and doing it line by line of code? Like, how should we be thinking about vibe coding in the world of 2025? Will, I'm gonna go to you first only because you know your time at Tesla. How many robot arms did you purchase in that timeframe? So let's just hear your take on this.
[00:03:52] Speaker A: That's a good question. I wish I kept an exact count, but I would have to say it was north of 5,000 robotic arms, probably closer to 10,000 or so. That was an incredible time to be there. I mean, we were building all types of manufacturing lines and I think it's underappreciated. Just like how much work goes into programming a robot to do a specific application. You know, in 2012, when I started in that role, this idea of collaborative robots or easy to teach robots was still very, very early.
Most robots back then, and even still most today are actually pretty dumb robotic arms. They will go to an XYZ coordinate in space and they will activate when you tell it to activate and deactivate when you tell it to deactivate. So grabbing or welding or whatever, there is an entire domain of work dedicated to people who are robot teachers and Robot programmers and the robots don't have eyes. They, for the most part, don't know where they are at any point in space or at any point in time. So it's all like you're programming it to just do a thing in an automated way. I think if I'm new to Vibe coding, so for one, I think this is just like a new concept to me. Blake is our resident Vibe coder within the organization.
And, you know, it's just really incredible to see what's possible today. Just like the ability to tell a robot, like, hey, reach out and grab that phone or that, that tire. You know, grab it and then put it on pedestal A, and then weld it at spot A, B, C and D. It's like we're in this age of AI where it's now not just affecting the software world, but also now the hardware world. So I think an exciting thing that we're going to see over the next few years are seen an evolution of robotics from being really dumb kind of blind movers of materials, to, like, you can grab a robotic arm and guide it to the points in space where you want it to be, to now just being completely hands off and just telling it what to do and making it happen.
So I'm excited to see the evolution of it, and I think it's going to happen faster than any of us realize.
[00:06:10] Speaker C: Greg, I got to hear what your perspective on this is. It doesn't have to be about programming even. It could just be more broadly about how you're thinking about how some of these concepts and ideas come to life.
As a former buyer and someone who lives and sleeps with capex on your brain.
[00:06:26] Speaker B: Yeah, I think, like, I have heard this example of that Vibe, that this coding, that coding is going to become this abstracted layer that sort of disappears and will just be, you know, interfacing with an end product immediately and using either natural language or other much more intuitive inputs, which, in some ways you theoretically have certain job functions going away. But like, in some ways, like, everybody's a programmer now, and I think that does port over to designing factories or manufacturing or buying equipment, and that you need so many specialties within these industries, whether it's just like every nut and bolt needs to be designed individually. And I think there's a good chance that AI enables systems engineering to be done in a much more effective and fast way that even teaching robotics, I think, will align to, like, the current best practices. Like, you can use a teaching pendant, but you can, like, hold it and move it it and guide it. But you see Tesla and a lot of other people using like YouTube videos where like the robot watches a YouTube video, abstracts what it is, and then does the thing in real time. And so there's no deterministic code, there's no teaching pendant, there's no other thing. And like you just go right to the result just based on watching YouTube videos. And I think the, you know, there is some job functions and things that change and I think how those paradigms of machine learning are applied to like other domains, which is like, currently we hear a lot of it in software coding, but I think you'll see it in mechanical engineering and process engineering and in electrical engineering. You'll see those AI applications work in almost any domain and certainly in any engineering domain as well.
[00:08:04] Speaker C: I don't want to get into how you think vibe coding will play out for CapEx and CapEx buyers, because I feel like there's some things you're probably cooking up in that regard.
But are you saying that vibe coding and more natural language interaction could start to play out for facilities managers, ops folks, other things as they're starting to think about the layout of their factory and essentially vibe coding, the way you think about layout of factory planning and that sort of thing?
[00:08:30] Speaker B: Yeah, well, maybe in some ways I just don't like vibe coding. There's too much in the zeitgeist, at least in like the tech world. I think the AI stuff is going to superpower a lot of people. You know, like Will was the first Capex buyer, Tesla, but they were a big, relatively big company then. You know, they were a public company without like Thousands or maybe 10,000 employees or something. And you need to be spending probably billions of dollars in capex or at least hundreds of millions to about buying like hiring specialty people like that. And I mean, I think what Diagon's building or what AI will do a lot of is like you will effectively have a Capex buyer. So any engineer. It's common that for smaller companies, the Engineer or the CapEx buyer, the. The CEO is the CEO, the CFO. And we'll be able to give the skills that say Will and I have to anybody or a mom and pop shop that's looking to buy their first CNC machine for a million dollars.
[00:09:25] Speaker C: All right, so hold on, I need to take a step back. The vibe coding zeitgeist as a whole, what is the vibe check there? Will and Greg, yay or nay. What do we think? Are we calling it vibe coding and manufacturing industrials or are we just ditching it all together and we've gotten overhyped in this.
[00:09:41] Speaker A: I mean, I think we can call it whatever we want, but, like, really, I think that the connotation that comes with Vibe coding is that it's just, like, kind of fast and loose.
And I think the real.
There's just a lot of hype behind that and the ability to just, like, build anything and do it from this abstracted layer where you can just speak it into existence that I think might be overhyped right now. But what I think is real, like the real takeaway that I have from it is that the results, or at least the end product that you get from a Vibe coded application or program is that you can get to a pretty quick prototype within minutes that's really good at demonstrating functionality. You know, how you want something to look and where you want it to pull information from.
All of that stuff I think is super useful. So the way that I'm thinking about it is, you know, when we think about developing software, it used to just be relegated to, you know, to figma, building up a demo there and just having, like, some more or less static. You could do some prototypey type of things in it with clicking of buttons and things like that. But actually being able to have, like a real tangible, you know, just like a prototype, a dumb prototype that you can play around with and give the vibe of what you're building. That is really impressive.
[00:11:01] Speaker C: All right, so it sounds like Vibe coding is a name. We're ditching it, we're getting rid of it. But what you can do with it, as you're thinking about these models and how it applies to, you know, both equipment procurement, manufacturing. That's got some legs. Sounds fair. All right, so I'm going to jump over to a topic that I feel like got Greg so fired up that he's even turned off his air conditioner in San Carlos right now just to make sure he remains fired up. And, Greg, the phrase I'm going to let you say it, drop it on.
[00:11:30] Speaker B: Us one more time, is that reshoring needs a rebrand.
[00:11:33] Speaker C: I feel like you need to say it with a little bit of gravitas there. Reshoring needs a rebrand, and he's been hammering this drum for a while. For anyone that listens, I would say half of his birthday party was just talking about reshoring needs a rebrand.
So, Greg, tell us what that means for people who are listening and maybe don't necessarily keep up with this idea of reshoring and what they're seeing in the news, what they might see on Twitter, LinkedIn, that sort of thing.
[00:11:57] Speaker B: Yeah. I think what I don't like about it is I at least get this image of like going back to the factories of the past. And I think Diagon's tagline is building the equipping the factories of the future. And like, I want it to be a more aspirational version of what the factories of the future should look like, that currently look like, and what they could and should look like in the future.
And I wanted to be a more compelling version. I wanted to be.
Yeah, I want to be the factories of the future and build these futuristic. I don't think they are lights out, nobody there. Like, that's not the. The future. The factories of the future. I imagine there's still people, highly skilled workers, but they're building strategic high technology items that the US or other countries need. And that's compelling to me. Bringing back the factories of the 1930s or 1950s is not compelling to me. And I think most people. And that at least is the image in my mind I get when I hear reshoring.
[00:12:49] Speaker C: I love that. So bringing back critical capabilities to the US using modern equipment, modern techniques, and ultimately upskilling and reskilling existing folks to bring those careers back to the US as well.
[00:13:02] Speaker B: Exactly.
[00:13:03] Speaker C: All right, so the one question I'm going to ask you then is, and you're thinking about one bit of advice that people listening at home are going to take away from this. What's the one piece of advice you want to give them if they're going to take this back into their operations, Whether they're a buyer, maybe an engineer, whatever it might be. One bit of advice to think about as they're trying to understand reshoring needs, a rebrand maybe to friend Shoring. I don't know. It sounds like something in that realm.
[00:13:28] Speaker B: Yeah, I mean, friend shoring is part of it. I think that's like, as any sourcing manager, especially if you're in production, like production procurement. I did a lot of too, where you're buying parts all the time, millions of parts per year. It's even more critical then because you're dependent on the same suppliers for many, many years in a row.
The economics of it matter like that. Near shoring and the unit economics, the return on investment.
I don't think people are going to nearshore for patriotic reasons or just like nostalgic reasons, like it has to make economic sense and whether the government has to provide subsidies where maybe it's not fully economic and there needs to be confidence long term that those subsidies or things will be there or it has to be fully independently has to be able to stand on its own two legs and compete globally. And I think that's what I hear from investors and other finance people that the risks of tariffs are like they make you non competitive. Right. They sort of, they create, they level the playing field but also they, they don't make you directly competitive with other countries but you really need to be like, you need to be globally competitive and, and I think automation and like this future of manufacturing is a huge part of how you, how to be globally competitive with places that have you know, one fifth the labor cost.
[00:14:43] Speaker A: Yeah. The reality is that you know, most people end up paying for the tariffs rather than making a real, the real decision. The real incentive is supposed to be that tariffs make it more economically feasible to reshore or build the things that you, that you want to build closer to home. But you know, the behavior that happens short term is that if it's not painful enough then what most people end up doing is like making the decision that they're going to just pay for it, absorb it and probably pass it along to the consumer, the end user or someone else in that value chain.
[00:15:16] Speaker B: Just shows up as inflation. Not as like this leveled, leveled playing field. It just shows up as increased prices.
[00:15:24] Speaker A: I've got a slightly different take Greg on reshoring. So I think you're right that in this era there's a fair amount of friend shoring where bringing manufacturing closer to the US bringing by, you know, building it in places like Mexico and Canada and other kind of friendly nations in Southeast Asia is part of the strategy I've seen companies employing. But I think there's, there's still a really powerful version of this that's bringing the manufacturing like bringing those factories to the US There was like a video clip that a bunch of my friends were sending around. It was Dave Chappelle talking about reshoring of manufacturing.
It was him essentially saying like I don't want to make Nikes, I want to wear them. So he's getting at is like, you know, there's this idea that by reshoring what we're trying to do is to bring you know, these low cost of labor jobs that are really like, I don't know, not, not high value add back to the US and have people in Nebraska making, making Nikes by hand. And I think it's just a very misconstrued idea of, of what this really can be and what it should look like.
The thing that I think about is like what things are critical for us to be making here. When I think about it, it's much more about capability building. Do we want to be able to build semiconductors if we need to? I think so. And is it a strategic industry? Is it something that can be automated? Is it high quality? Is it related to national security? I think all of those things, it checks all of those boxes. I think the same thing for aerospace and defense products. I think the same thing for energy products, things like batteries and nuclear reactors.
So I think we've got to just kind of uplevel the way that we think about which industries are going to benefit, which capabilities we need to have and want to have in the event that we have to make some of these things on our own. And then the last part is really around the labor force and what this looks like.
I think that there are a lot of people who turn their noses up at factory jobs. But the reality of most minimum wage or low to kind of low wage jobs in the US is that they're mostly quick serve restaurants. So think like Arby's or Taco Bell fast food restaurants and things like that. Entry level jobs at places like Target and Walmart where you're a cashier, you're stocking shelves, you're you know, doing something like that or their Amazon labor, Amazon operations and people who are running these big operations and distribution centers and you know, not to knock any of those jobs, I think that they're, you know, if you've got to do the job and you know there's a need for it, then do it. But if you were to ask me like would I rather work in a semiconductor plant where I'm learning some skill as an operator, contributing to a product that's going to go into something that's high value add downstream. I've got a career path that's impressive and you know, I'm able to work my way into higher and higher value jobs. Like that for me is the whole game. We shouldn't be thinking about this like dark, dirty, dangerous factory, the place that you want to work. But really thinking about like what are your alternatives? Like would you rather be working in a shoe store or in a factory where you could be making some really cool products?
[00:18:43] Speaker C: All right, so what you're saying is capability building for semiconductors is conducive to careers. What do you think, Greg? Capabilities. You agree? Yes. No. What's your take?
[00:18:53] Speaker B: I do think there is a Cultural aspect of, and I think President Trump hits on this, that the deindustrialization just gutted the blue collar Midwest worker. Where you could, the American dream a lot of the times was to get a good job working in a factory. I mean it's sort of romantic, but I think it was true. And you could, even on a single income, if you got a good job with a good company, you could work in a union for 30 years and you know, go to work. You worked hard five, five or six days a week.
And sometimes company provided housing and benefits and people built their whole lives around that. And I think that dream is not fully dead, but it's like it's not as easy to attain. There's a lot and there's lots of dynamic globalism. Part of it, the unions are part of why it's, it's not what it used to be, but that's real. And I think like globalization has benefited a lot of people. And honestly people like us, like educated coast people in general have benefited a lot. Like corporations broadly have benefited. But I think a lot of blue collar folks that like we just described like that dream is like not attainable or you have to go, you know, to do other things that are just not as attractive. And so I think that's part of the romanticism. People want that dream to be alive again. But I think it is harder. It's like the, the lower skill jobs that sort of anyone could do probably don't exist as readily. And so you need some level of training, some level of expertise and skill. And I think that that's not a trivial problem. And I think the US is not this the US Canada and other countries have dropped the ball on like what retraining looks like. How do you upskill people? Because it happened pretty fast. Like it was over, it was in less than a generation that these factories went from these vibrant towns to you know, hollowed out places. It's, it's incredible to see it in person actually.
[00:20:44] Speaker A: I think that the perception problem about manufacturing is that these are low skilled jobs and like why would anyone want to do them? But the reality is actually that some of these jobs might be more high skilled requirement than what most people are trained to do. So if you're going to work in a semiconductor plant, the primary concern that these companies have is like, am I going to find a workforce that's interested enough and educated enough to be able to operate the equipment, to identify defects, to manage exceptions when they happen?
And so I actually think that the opposite is true. Like we're going to have to do some amount of training to make sure that people are ready for these jobs when they do come back.
[00:21:25] Speaker C: So quick lightning round question then for you both. We just talked about jobs and skills and capabilities. What are these top one or two skills or capabilities that you think individuals at home should be thinking about or telling that family member who says wait, what should I be telling my kids to do when it comes to manufacturing and industrials?
What's like the one or two things. And I'm going to keep you both on the clock on this one so you can't wax poetic about too many skills and capabilities.
[00:21:51] Speaker B: I mean I think any skilled trade, any skilled certified trade is, is very attractive and I think if you work hard there's a lot of opportunity in these industrial centers right now. And I think and getting closer to the value which is like, I mean it's probably more skilled trade but like controls, robotics, automation is likely going to be higher paid, higher opportunity, but it's, it's obviously like probably a higher bar for requirements like if you're trained and if you're trained in any skilled trade and certified, there's a lot of opportunity for anyone willing to work hard.
[00:22:24] Speaker A: Yeah, that's not the direction I thought you were going to go. I thought you were going to go for. I'll maybe pick the opposite side of the like once you've built the factory, what I thought you were going to say are the jobs inside these factories of the future. And so maybe I'll mention that I think that machine tending and machine operating is going to be a really important skill. You know, there is a lot of talk about lights out factories and you know, these factories that just kind of run autonomously. But the truth of the matter is there's going to be a lot of root causing troubleshooting and trying to understand, you know, really still having an understanding of what the machines are doing and how to get them to correct when there is a defect issue and things like that. I think that's going to be something that's, that's really important. And I also think that like maintenance, repair and repair operations is going to be a big thing because when things do go down I think it's going to be a really important to get them up quickly. So understanding, you know, still how to repair and how to keep these machines running in top shape, that's going to be something that's really important.
[00:23:28] Speaker B: I talked to a company that trains people to run coating lines like that coat material onto to like to metal, like say foil. Like almost like aluminum foil.
They said it takes almost two years, like, and you had it's on the job training, which is incredible. But here's like a thing where if you have a defect, the coating machine codes almost a football size field piece of aluminum every minute. And so like if you have a defect that's running for five minutes, let alone like an hour or five hours, like you're printing like a football, football field length at least, and these things are probably a meter wide or maybe more, that is just trash. And then that sounds bad enough, but if that foil ends up and the defect, you may not even see it with the naked eye, which is also crazy. So like you need special computers or cameras to see them.
It's even worse if that foil gets built into a battery and then you don't find out for like a week later. All of a sudden you have a whole factory or a whole warehouse full of batteries that don't work.
[00:24:29] Speaker A: Or God forbid you've got, they've made it into the market and you've got to do a recall.
[00:24:33] Speaker B: Yeah, or a recall. And so it could be just this guy, we'll say a guy or girl that runs this coding line that doesn't seem like a high skill job. But if they are not paying attention for even a minute, let alone hours, or one of these mistakes goes unseen for a long period of time, the consequences of failure, and there's some automation built into them, but there's still a lot of human judgment that goes into running these machines really effectively.
[00:24:58] Speaker C: All right, so what do we think then? Lights out factories. I feel like these have been in the zeitgeist for at least a decade, if not longer.
Do we think it's the false prophets of past are still false prophets when it comes to this, or do we think we're coming up on it soon? It sounds like both of you are not super bullish. And we need people tending the machines. We need people in the mix to feed the machines.
What do you think? Fair to say?
[00:25:22] Speaker A: Yeah, there's always someone that's interacting with the machine or getting equipment, getting materials from place to place. I mean, we'll slowly be eating our way into automating, you know, the operations and things related to it. But I think we still got a ways to go.
[00:25:39] Speaker B: Yeah. I used to think even Tesla had the most advanced factory car factory in the world. And Will and I used to go to work and there's 10,000 people that went there every day. And I think even if Optimists like, I think about what the jobs that that Optimus could do there. I'm confident there would still be like 6,000 people that would go there every day, you know, or, and maybe the productivity of the factory could be higher, but there's still a lot of jobs and skills required, even in the most advanced factories in the world.
[00:26:05] Speaker C: You know, that's a good transition to talking about factories. I feel like we're seeing a lot of these factories, both new factories being announced every other day, whether it's because of additional pushes from the new administration or just in general in response to various types of tariffs. But we're also seeing the reinvigoration or refurbishing of existing factories.
So I'm going to come to you in just a second, Will, because I want you to explore, explain to everyone listening at home, why would you build a new factory versus why would you refurbish or reinvigorate an existing factory? But before I do, Greg, buy the numbers just because I just know you love some good numbers.
Just this last week, Emirates Global Aluminum selected Oklahoma for a $4 billion investment to have a new aluminum production facility, which is the first of its kind since the 80s. Literally since the 80s. 350 acres of aluminum. So going back to the kilometers versus miles you mentioned before in football field sizes, literally, we're talking about a massive, massive facility. Even the GM Buffalo plant, GM announced, I think it's their engines, right? I think the new engine facility. 888 million. I'm not sure why they chose 38 in Buffalo, New York, basically retrofitting and building some new production lines. I've got to ask, ask to start before I go to Will to explain both of these topics.
Why do we have new factories? Why are we thinking about new factories in general?
What is spurring some of these new factories and the idea behind it.
[00:27:37] Speaker B: I used to source aluminum for Tesla, so I have some experience. I've been in aluminum factories like this and steel, both steel and aluminum ones. There's a big dependence on China for aluminum. Right? And I think you think about steel. Aluminum tariffs have been in the news for a long time and they have a crazy amount of capacity, like batteries and others. They have a lot of capability. But aluminum for aerospace, there's a lot. I mean, steel and aluminum, like makes everything, you know, ag concrete. It probably is like the three most important products in everything we make.
And so it's the base material that makes everything. It's also like highly recyclable. It's like the most recyclable metal. And so it's probably there's a very high demand for domestic aluminum. And if you can do recycle, I don't know what this specific facility is. The inner. The one thing I learned about aluminum, which I thought was really cool is like to make aluminum from raw material versus recycled material. It's like 1/10 the energy to remelt recycled aluminum. And so if you can do like Ford and Tesla was working on this like what they call closed loop programs where you basically take the old aluminum or scrapped aluminum from your old cars, like the Ford F150 and a lot of the Tesla cars have a lot of aluminum in them.
You basically take them back and you have this closed loop system where you're using the scrap or recycled material and you can recycle aluminum effectively infinitely. And so it's super. It's way more energy efficient and you can use the material over and over again.
[00:29:03] Speaker C: Will, what do you think? Is this a new capability? Is this us leaning back into building some of these capabilities? Going back to that reshoring conversation as we think about aluminum or if you're in the UK listening to this, it's aluminium, I believe. Will, what's your take?
[00:29:18] Speaker A: I think it's really all about proximity of the supply chain. So there are some plays that are really strategic because there's a thing that we want to build and it has to be built domestically. And then in other announcements where you're talking about something like aluminum as a raw material or battery materials and things like that, it's all about domesticating the supply chain. So having a domestic source that you can go to.
So aluminum manufacturing isn't new by any means. Like, you know, my hometown Pittsburgh was like the home of Alcoa and we had US Steel that was also like a massive steel fabricator there.
So, you know, the capability is not new. It's been outsourced for a long time. But as we start to see more automotive battery and especially drones, like a lot of the flight vehicles require this type of aluminum. It's really just all about creating and having domestic sources where we can build them.
[00:30:16] Speaker C: So Greg, you have experience sourcing some of these things in terms of actually making facilities just like this. What is your take on where the equipment that's going to fill this factory, where is that going to come from? Is it going to be all new? Is it all made in the us where does that type of equipment come from?
[00:30:33] Speaker B: I think we talked about it. I haven't bought aluminum equipment. Before I was buying actual aluminum that would chip into Tesla. But. But I'm very high confidence it comes from overseas somewhere. I mean, there's all across the board, there's very little equipment made in the US and so 60 or 70% of the equipment that likely is going into that factory will come from overseas somewhere. From Western Europe or from Asia likely.
[00:30:54] Speaker C: All right, so Will, I'm going to go to you then. As we think about new facilities versus refurbished, reinvigorated facilities, explain to people at home, and I'm going to put it on the clock here under 90 seconds.
Tell me why someone would build a new facility versus why someone would refurbish or reinvigorate an existing facility.
[00:31:15] Speaker A: If you're going to build a new factory, you've got a lot of things you need to think about. One, you need to select the equipment that you're going to put into this space. And even preparing the building for that stuff could take a long time. You've got construction considerations and timelines that you've got to work around.
You've got to get utilities run to the building. You also need to think about permitting and requirements, so getting permits for air and water for that location. So there's just a lot of things that you'd need to consider. The benefit is that you get much more control over what you're building, but the timeline can be longer and it can end up costing you a bit more in the end. Where, like if you were to go refurbish, you could, let's say, buy a building that was already built for that use case as its intent. Then you can get grandfathered in for certain types of air permits and things like that. You may also benefit from any existing infrastructure or equipment that's there. And usually you can shave months, if not years off of the timeline.
So it really just depends on the exact use case for the customer. But there are benefits and drawbacks to each.
[00:32:24] Speaker C: I love that it's a good use case of used versus new, honestly and thinking about as you're trying to break that down. So. Thanks, Will. All right, Greg, I'm gonna go to you. New versus retrofit. What's your take? Any, any hot takes on this or perspectives that you'd add or disagree with.
[00:32:39] Speaker B: On Will, the hot take is boring. It's like it just depends. It depends on the customer. If you're a new startup like you probably speed is most important and you can probably make it work with existing facilities if you're a mature company that knows exactly what you Want, you have the capital and you kind of have the time, you kind of want to go greenfield. That's kind of what Tesla did. Is it scaled? You know, the first factory was a brownfield existing facility. It gave them like, I think it's the second biggest building in the country, able to scale. They could fit three full model lines in there, kind of four. And so that gave them a lot of flexibility to move quickly where, you know, as they've scaled out, then they've started to build greenfields in China, in Nevada and Austin and Berlin. And I think it's not always the case, but something like that work would probably work for most people.
[00:33:29] Speaker A: Yeah, I think I'm pretty aligned with Greg on that. And just thinking through, like, what's the stage of the company that kind of determines which things you're going to be more open to?
You know, we talked about these, the old days back when we were working at Tesla, and I just remember what it was like actually inhabiting that space.
Maybe just to go on story time for a little bit when I moved into the factory, imagine this is like a. It's a five million square foot facility. It's massive. The main building is like two and a half million square feet. And then you've got a bunch of ancillary buildings that house like plastic injection molding and painting and all that stuff. But most of the factory was completely dark. Like it was just really cavernous. We had a small operation in the back where we just cleared out some space. All we needed was factory floor space to assemble the vehicles. We did most of it manually. And you'd walk past like line after line after line of old Toyota equipment.
And, you know, even in some of the places, like we did a refurbishment of the north paint Shop while I was there, the North Paint Shop and the South Paint shop. And it was surreal. Like, we went into the North Paint Shop when we were like just tearing out all the old equipment. You know, you're going through lockers that have people still have their stuff in there from like the 90s. You're finding people's lunch boxes. We found like a dog carcass in one of the buildings too. Just this, like animals and, you know, hideouts of people who were using that space for the last decade.
And the one requirement that was imposed on us by the city of Fremont was that we couldn't tear down any external walls, otherwise it was going to be considered to be a new build. So Elon was maniacal about that. He's like, don't touch Any of the walls you can do a tear out is what we called it. Not a knockdown and rebuild, but we literally just like, opened up one of the garage doors and we're stripping out all this old conveyor equipment.
The benefit that came from that is that we didn't have to repermit as a new paint shop. It was already zoned as a paint shop. We were able to utilize some of the same ventilation equipment and other things that were already in place.
And there's also a lot of environmental liability that comes with owning and operating a place like that. So Toyota never actually shut down their entire operations, because if they did, it would have meant that they were responsible for starting to do the environmental cleanup. And so they had, like, a skeleton crew that they kept there just to be able to continuously inhabit that place until they found a new buyer. So it was really. I mean, it was a. It was a fascinating case study in how to refurbish a factory. I see a lot of companies now taking advantage of similar strategy. I was at Astra more recently, where we were building a rocket factory inside of an old jet engine manufacturing plant in Alameda and similar type of deal there. It was already permitted and entitled as the kind of place that it was. So you get a lot of Runway out of these old places. And I think it's. I also just think it's cooler to hang on to some of the history that comes with the building.
So if you can preserve that heritage and bring some jobs back, then you're winning.
[00:36:40] Speaker C: I love that. Well, I think that's a great transition, too, of thinking about how retrofitting versus new. I mean, there's an opportunity to also add more legacy to a place that already has a lot of historical legacy. Like you just mentioned, you playing this part, filling the factory for the new Tesla factory that was the former Toyota facility in Fremont. All right, gentlemen, let's go ahead and transition to. What is my favorite part, and I think your least favorite, the hot takes part of the Movers and Makers podcast. Will, I'm going to you first. What are some of your hot takes?
[00:37:09] Speaker A: My hot take is that we need to hear more from you. Blake. One thing that I didn't realize about you kind of expecting that, you know, you're this. This marketing executive, you've got a lot of great experience, but I had no idea how many AI tools you just have in your repertoire. Like, we were just on a jam session the other day, and you pulled up at least 10 tabs of, like, 10 different, like, an AI tool for this and a tool for that.
So I want to hear more from you on that because I feel like I learned something new every time you open up a browser. And you know, the ethos that I have for this is Greg actually will appreciate this.
He's gotten me to be a big fan of Marcus Brownlee, the guy that does these reviews on cool high tech products. I feel like I need Blake's takes on these that we should coin that Blake's takes on the AI software because when someone like Sam Altman tells me, hey, here's this cool new tool that you should try it, I feel like the barrier to entry, like the intelligence level that I need is way up here where like, no offense, dude, but like, when you tell me, I'm like, all right, if Blake knows how to use this, then I can probably learn it.
So yeah, I would love to learn a little bit more. And I feel like I'm already up in my game on what tools are available and what it does.
[00:38:26] Speaker C: Like, Blake's takes is going to translate into a branch off of this. I mean, I will say, and it's worth mentioning as we're talking about all things manufacturing and industrials right now, the things that are happening within AI are exciting because people are taking the first phase, taking unstructured data, data that has been tagged and metadata, et cetera. And then now we're moving to this next phase of using AI to turn the unstructured data into structured data.
And as we do this, it is truly infiltrating and getting into every industry and vertical that we can think of. Salesforce just bought a company for almost a billion dollars that it was a metadata company that would have never happened 15 years ago. And that is all in the service, structured and unstructured. So I think for the things that I'm excited about are as we think about manufacturing and equipment and building the factories of the future, whether new or retrofitting, is how are we using these AI tools for good and making more sense of the structured and unstructured data that's out there and translating it into usable data for all. So I'm pretty excited about it in the context too.
[00:39:36] Speaker B: I mean, Diagon's obviously working, cooking up some cool things, but there's just some part of it is for regular people that are just like working in jobs like we all are, or bigger companies, like, I think even just giving some insight into like how what tools they should be using, experimenting with whether in their personal life, but to make their jobs better, you know, I started using chat GPT to help write emails, but now it's like to analyze data or to do. I was trying to get an estimate yesterday on how much it would cost to like carve out a big concrete slab to fit a big machine in and use deep research. And it took like five minutes to go and do all these assessments. And it came back what felt like a pretty good thing. And so even just getting some of those tools onto some listeners radar and even us brainstorming with each other like it's like the, the age of the Internet. We've talked about when it's like there's a new website every day, there's like a new AI tool and some are stupid and pointless, but there's a lot of really powerful ones too. And it's great to try and separate the wheat from the chaff, I think is the term.
[00:40:36] Speaker C: I think we've got the way to think about AI is we've got the smartest intern that ever was or ever will be that has never stepped outside of the office. And they're relying on us to make sense of all this industry expertise.
And so I think how you would use that smartest intern ever and feed your industry expertise to it is the art and the science, which we will not call vibe coding at all. I think we rely on that. So that's totally fair.
[00:41:07] Speaker A: So here's a question, Blake, maybe pulling the judo move for you on the hot takes. What's maybe your AI Blake take of this week?
[00:41:16] Speaker C: So my hot take of the week Will, is I'm going to say that right now we're moving from structured text based data in a lot of the ways that we use ChatGPT or Claude or Gemini or any of these open source models to we are moving quickly into 3D data and not just 3D data as a standalone, but we're moving to 3D data with additional metadata, which means we are moving into a world where truly the limits are expanding of what you can do with both media and entertainment, crafting movies and games of the future. But as it applies to manufacturing folks that are listening right now, the way that we interface with CAD files, the way that we interact with 3D files and in the future 4D data. Even as we think about LIDAR data that comes off of things like the cruise cars of the past and the Waymo cars of the future, all of that is structured data. And these models are getting better at handling massive quantities of it and querying it. So I think my hot take exciting thing of the Week just that I've been reflecting on is how can we expand our horizons into 3D and 4D data and apply some of the same product primitives that we've been learning from this text based interaction into some of these verticals. And I think some folks are cooking up some really exciting things in that space that I'm excited about.
But that is probably my, my hot take or thought of the week I would say.
[00:42:38] Speaker B: So I've seen that firsthand actually. I've been advising a company, Collab Software for the past seven years and they do, they have like a tool for mechanical engineering collaboration so you can kind of do design reviews in their tool. But they're obviously starting to implement like AI review it's called. And what they're seeing is like originally AI could say, oh, this radius is too big or you may have this like mechanical design property. But they upload it like a PCB design recently that had like a circuit board with a fan on top and then it started to return things to say, well, we think the fan could create mechanical issues. So like it understood like the operation of the board and like potential issues with like heat management. And it had very little context on it besides just like a step file. And so it like it really like the LLM understands the context of it and it's like, wow, they didn't even expect it. They're six months ahead of where they thought they would be with the AI tool. And just the context and intelligence that these models understand is incredible.
[00:43:43] Speaker C: It's pretty amazing. Yeah, I think.
Well, I don't want to reveal too much. Diagon's cooking up some exciting things.
So for all of you listening, tune in for next week, movers and makers, as we dive into a couple of these additional topics. Any other final thoughts, Will or Greg, on all the things that we covered today or that we didn't cover today that we think it's important to share with folks that are listening today.
[00:44:04] Speaker B: I said to the team recently what a time to be alive. I think between AI reshoring wave, I always had this FOMO of not being in Silicon Valley during the Internet wave and I think not to be. I think this is a better time. I think it's going to be bigger, more impactful and then I do think there's going to be this reshoring wave. It's going to look different. But I think two of those things combined together, it's a pretty incredible time to be building a business in both those spaces at the same time.
[00:44:30] Speaker A: Yeah, I agree My final thoughts on that are I think I've been surprised by what AI will tell you if you ask it.
So I think that there's this kind of behavioral pattern with my generation where we just kind of take the results of a search as at face value, like, okay, here's what I asked Google, here's what it returned me. But you can't really ask Google, why did you show me this?
It's not really built to do it that way. The really cool thing about AI that I've just been experiencing is like, you can ask it like, hey, why did you return this? Or like, hey, I didn't get this same result. Like, where did you see this information?
Or like, you can have it check itself, like, hey, you gave me this result. You know, can. Let's now build in a second step where you go in and peer review this. It's really cool to see how, you know, you can kind of split the personality and cross examine whatever comes your way.
[00:45:28] Speaker C: Truly, as Greg said, what a time to be alive. Gentlemen, thank you so much. For everyone that's listening at home, thank you so much for tuning in to this week's episode of Movers and Makers. If you have any questions, comments, feedback or thoughts of your own about what we covered in today's episode, reach out to one of us on LinkedIn. Send us a DM, leave a comment if you haven't already. Be sure to leave a good review on Spotify, Apple Music, YouTube, and be sure to like, follow and subscribe. Gentlemen, thank you so much. I'll see you next time.
