Hey everybody, Tim from Alfolf Capital coming at you with a follow-up interview to one of my favorite absolute favorite companies and CEOs, Allan Evans from Unusual Machines, ticker symbol. Allan,
I'm good. It's been a little while. It has. It has been a little while. And you are never boring.
I don't, I guess sometimes boring's good.
Yeah, maybe for somebody. But I actually, you know, and I said this to you before we hopped on right now, is that you never cease to amaze me as to where you get involved in things and where you see things going. So, I think you had a really interesting announcement with your investment in Extend. I thought that was really an interesting thing. And part of the reason that is so interesting for me is because I've just recently learned about the whole physical AI thing, right, through Tyler Gates over at Bright Line. And he has opened my eyes to so many possibilities and things that I'm excited about the future. And when I saw that you were with Extend, can you maybe just share with everybody what Extend is, what they do?
Yeah, sure. Extend is an Israeli drone company working on what is XOS, which they're really interested in software to help control and integrate information from a bunch of different drones. So again, you know, a lot of the work in Israel, a long time building a bunch of drones. You saw that one of their drones was selected for the gauntlet program for indoor for instance. So they're still very much doing that, but they're really really interested in sort of both the onboard vehicle and then the multi-vehicle control layers in terms of sort of using agents to help control them and they're not so married to just their own products. Like they're interested ultimately to expanding it to as many hardware platforms as possible. So, I mean, from our end, they're building drones. They're winning awards. They're important right now. And we invest in a lot of our customers to help them bridge the time between when they have to buy stuff, to build stuff, to ship stuff. You know, we help with that flywheel. It's like we take a little equity for credit terms. You know, that's how you could think of it. It's pretty simple transaction from our end in customers we believe in. And then I think longterm what they're doing is really cool in that they're trying to help build this abstraction layer to help control a lot of different vehicles,
Right. So it, you know, that's exactly I guess the best way to describe it is it's an additional layer and it's hardware and the chips and it's a different layer. It's kind of like a toll road, right? Essentially,
I wouldn't look at it like a toll road. Like physical AI feels really jargony to me and I try like if I can't explain what I'm talking about to my mom, I'm using words that don't make any sense to normal people.
Look, the how I see physical AI, my interpretation is it's basically AI meets sensors and robots. Tesla is a form of physical AI.
Right. It's taken in all this real world data. It's using compute. It's using an agent there to make determinations and then it's using the actuators of the car to change how it goes. You know, a lot of what we see for humanoid robots out of figure and stuff, these are agentic based forms of physical AI confined to a single robot. And what I think when you talk about Tyler at Bright Line or XOS or some of these other things are where I think people are really excited and use the term physical AI is where these robots start to coordinate past just the individual things. So if a fleet of Tesla cars were all talking to each other to optimize traffic flow, that's where people are using physical AI. So and if they were working with sort of cameras along the road. So these different systems, the running agents, it's a framework for these agents to talk to each other to get coordinated output. But that's the same as think of a rowboat, right? If you and I sat in a rowboat and we were not talking to each other, communicating at all, the boat would probably spin in circles, you know, because we wouldn't be rowing together. And so what physical AI is is the coxswain in front of a crew boat being sure all these individual pieces are all rowing together to get the most out of it. And so I think you're seeing these very early groups of people trying to do it and they're really trying to do it for the first time. I think if you look at Extend or Bright Line as opposed to where you might see Tesla or some of the others, they're trying to do it for robots made by different people and groups of sensors across different entities and trying to create this way for, you know, people on different teams to all communicate. So instead of, you know, I feel like Tesla's like everybody speaking Spanish and then people get in and they're speaking German, right? And what the sort of XOS is or what Bright Line's trying to do is sort of get everybody to use the universal language. They're basically Google Translate to the robots,
You know, and that's how I see what they're trying to do. I find that as a long-term future very interesting, but more near-term, look, they're building drones. The drones work really well because they're doing AI things. They're going to need parts to make the drones work. And it's really only at scale when there's a large number of actual things out there where you start to understand how to solve the last 10%. And so we're really interested like we look at Unusual Machines. We're part of the robotic infrastructure like you can only build robots as fast as we can provide parts.
And you know so if we build out that infrastructure they need to get to scale early to understand what it is. Right. Like Tesla's physical AI self-driving is much much better because there's a whole lot of it with a whole lot of use cases.
And so the first step to really making these ideas good are to expand the number of robots and expand what's out there. And the rate of expansion is limited by the production infrastructure domestically. And good or bad, we're part of that infrastructure as Unusual Machines for some of the parts we do. And so if you look at like Nvidia, they actually can scale a lot faster because TSMC is their infrastructure. The fabs are their infrastructure. They design these things that are then go manufactured on the infrastructure. So we from the robotics industry or the physical AI industry, a lot of big forward-looking crap in there. We look more like TSMC as a limit to how fast it can go based on how much we can make and the quality of what we can make.
No, we were going to just jump right into the deep end here. Hello audience. If you were here to hear about drone parts, you just heard about the general attitude on the next 10 years of robotics.
You know, one of the things that is so amazing to me is, and I told you this before we hopped on, I mean, I met Tyler and I got so immersed in this physical AI thing, and I said, 'Hey, do you happen to know Allan Evans because I'm going to have to introduce you.' And he's like, 'Oh, no. I know Allan.' I was like, 'Why does that not surprise me?' Right? Because you are very tuned in to where things are not where we are but where Wayne Gretzky had the saying you skate to where not where the puck is but where the puck is going to be right and such a true statement in life in general you are definitely one of those guys that's skating to where the puck is going to be.
I try. At the same time I do think it's really important to demonstrate success along the way and get to market and get product out there because that is a huge part of the feedback loop that lets you go faster. So I think it's one thing to be like hey look in 10 years physical AI agentic robotics world peace all very likely but if you don't engage the customer at each step along the way and you don't get that feedback loop and you don't take your knocks and you don't make the improvements or solve the details. Whoever is going to solve those details even if they don't look forward is going to beat you there.
Right. I think this is one of the most interesting things about how I look at SpaceX. I think SpaceX has done a very good job of a long time ago being like, 'We want to go to Mars.'
Well, we're going to start by putting satellites in space. Oh, okay. They monetized. They got feedback. Rockets blew up. They got better. Okay. Well, now we're going to do Starlink. Oh, okay. Here's an engine of monetization. Here's we haven't gone to Mars yet, but SpaceX has made money and been able to learn and accelerate because of the feedback loops along the way. And so I think even if I can say, hey, look, I think agentic robots, which is just what self-driving cars are, I mean, I'm just using generifying jargon, smart robots that do different things, talk to each other inside a framework. Great. If I believe all that's true, you still need a physical layer to build the components.
All the way from the motors to whatever else and that's where we sit. But if we're not just building those parts for today, somebody who was without ever bothering to think about the future would actually win because people would use that infrastructure built on top of it. Right? The internet spread everywhere because of dial-up. People didn't go, 'Oh, we need cable or fiber.' They said, 'Well, we got these phone lines everywhere.' Great. Let's have computers talk instead of people. So, I think it's important to recognize that there is some benefit to understanding where you're going because you can make some decisions that enable it a little faster. But if you're not doing right now, and I think that's what our team does a very good job of is scaling it and meeting the needs of right now, then you don't fit this feedback loop actually lets you win,
I read an article somewhere, I can't remember where it was, but it was talking about supply chain for the Department of Defense and that specifically when it came to drones that not necessarily having the best all the cool bells and whistles and you know the latest and greatest whatever. What the real issue was is for this military is to be able to recognize, adapt and scale and then supply, right? I mean, so you can have all these great concepts or great, but if you can't produce it for five years, it's no good to them, right? They need you to be able to recognize, adapt, scale, and supply. That's it.
And that's scale. Scale is a big one.
It's a big one, right? And you're I would say experiencing that better than maybe anybody right now.
Yeah. Team's doing a pretty good job. Like we're building out our supply chains. We're scaling. I mean, you saw last quarter was a $16 million quarter. A year ago was 2 million. So we 8x'd. That's painful. You know, publicly said we're actually spending a moment this quarter to sort of tool to continue scaling forward. So, I wouldn't expect the same revenue, but then we're the infrastructure around which a lot of these programs depend upon. And so you know different drone companies could go from 1 to 4x if the infrastructure is there if the different manufacturing entities are there you know but we're one of those pieces of the infrastructure. So the industry can only scale as fast we or the other pieces of the infrastructure go. How many motors are there to go to the different factories that require motors?
Oh, you know, how many batteries are there that are available that meet the FCC and the NDA and the Department of War requirements? And so when you look, if you're starting a drone company, you could go find a contract manufacturer out there who could do the electronics and do the final assembly of the drone. There are a lot of them, you know, a lot out of in Michigan and other places from automotive that have retooled and figured out how to do this. And so you could start a drone company, design a drone, go do new product introduction with them, and use just some of their utilization to very quickly 10x your production rate, but your production rate is actually limited by your supply chain, how much stuff you can get in. And you see this with the cost of memory on things.
But memory is a limit. And so that's a limiting factor from the infrastructure perspective because memory is one of the technology infrastructure plants and you can't just infinitely have more memory and so you're seeing all these prices go up in the same way a lot of what we build is at the infrastructure level and so as fast as the domestic industry is growing and if you look at Gauntlet that's still running on time that's huge demand you look at Power US with the interceptor drones and some of the others that's additional demand like we have to continue to scale as fast as possible to provide this infrastructure that a lot of these companies can rely upon and it's really interesting but it's sometimes it's a painful mandate.
Oh, absolutely right. I could see that. I mean, you highlighted in the last quarter some of the challenges that you got had, but that also is what gives you confidence that you know, you have the right team, right?
No question. The team's amazing. And like some of those challenges obviously have ripples into this quarter, right? You have a motor challenge like that. You go, you ship out. Well, you still got to go replace everybody's motors, right? You're not going to let them put bad motors on products. So now, you know, you look in quarter three, like some of our revenue that we would normally see from motor production is just a zero as we take in some of that and scrap it. Like it didn't all happen in a single quarter. It's stuff like that. And then dealing with that while then also putting in the high-speed production line, clothing, acquisition, upgrade, energy. You know, a lot of these pieces of work are restructuring the company to continue to build infrastructure going forward. But, you know, that's part of where we have to navigate the whole thing. And every member of our team pretty much has autonomy to go do it because there's no time to babysit anyone. It's like your job, your job, your job, your job, and your job. Brand,
You know? I feel like I'm a huge fan of the NFL. I feel like it's an NFL game. If any one of the 11 people fails at their job, something bad happens, right? You know, if a quarterback doesn't cover the touchdown, easy. You know, an O lineman doesn't block, the quarterback gets sacked. So, I think it's very much everybody knows the play. Well, right, trust everybody to execute, but you can't do their job for them, right? And so, you know, that's what running our business looks like.
And I think you're doing a fantastic job. I mean, I actually So, you're getting close to 300 employees now.
Yeah, we're 280 not counting the employees at Upgrade Energy. So after we close that, we'll be above 300.
I just think back to when we first met and so Hey, when do you think you'll get the like what's full employment look like? 40, 48,
Right? I mean, but then again, you always said, you know, look, the storm, the wave, the wave is coming. You just didn't realize it was a tidal wave.
No. And now, now you pretty much got to surf it, get smoked by it. Like, this isn't one where you can just be like, I wiped out. It's comfortable. You're like, nope.
Nope. We're riding this one. Why? Cuz if we don't, it'll swallow us. Yeah, but that's a great place to be in, right?
It's exhilarating. You know, in the same way surfing a tidal wave, which I've never done, by the way, is probably exhilarating. You know, there it definitely hones you all the way to an edge where you're very dialed in and very focused on what you're doing in the immediacy and you know, have a chance to do something. You just don't get too many tidal waves in life.
I think a lot of really positive waves happening right now when I look at this. I mean, it's not just the this getting back to manufacturing in the United States, I think is just awesome. It in so many ways. I think it spurs innovation. I think I just think it's a really cool time. And I know people probably listen to me saying that. They go, 'Dude, like there's so much in the world.' It's like, no, man. I'm telling you, I got a really good feeling this is all going to work out just fine. I really feel that way, you know?
Maybe it's too optimistic, but I don't know. I wake up every day thinking that today's probably the best day in the history of the world. And tomorrow will probably be better. And I mean, if you look back over time, that tends to be true. So instead of thinking about oh no doom and gloom, it's like how do we take these tools to make tomorrow a little better? How do we make things a little nicer than today? You know, and I think to me that approach is a really helpful mental framework to not just get caught in this cycle of ignoring innovation because it's fear. Like it doesn't mean to say that there's not risks to anything or there shouldn't be concerns. Like I don't want to be dismissive of that. But most of the time any tool can be used for good or bad.
But we shouldn't be afraid of the tools and when they're used for good it's great. So even I had somebody ask me what do I think about this regionalization of technology this need for new minds in the US this onshoring to manufacturing and what I think about from a global economic long-term perspective. 40 years of globalization allowed us as a world to get the most efficient you know we had specialization like TSMC and Taiwan or the rise of excellent manufacturing in China and the movement of a billion people from abject poverty into the global middle class. Okay, that's pretty cool. Now, a lot of them weren't in the US. Fine. But now this recentralization, Europe's notion of tech sovereignty, the US re-industrialization after critical minerals efforts, what that's doing is it's taking this amazing world that we built through maximum efficiency over the last 40 years, and we're locking in redundancy. So we're not maybe moving ahead as fast as if we continued with globalization. We're not maybe doing that. But what we are is we're securing the gains of the last 40 years by having it not depend on a single location. You know, if there's world-class chip facilities in Arizona and in Europe and in Taiwan and in China, if there's a problem in any one of those regions, it doesn't matter. The world still has the benefit of this distributed world-class excellence. And so I don't even see it as a downside from an economic perspective because I think the gains of the last 40 years have been incredible. We're talking on the internet. We have these magical glass screens that can take pictures of everything. We have food systems where people in the Midwest can eat fresh sushi or right like that's all really great stuff. Now we're locking it in. We literally have a computer in our hand, right?
And a camera and a super high-powered radio that can connect to anywhere in the world. And an AI voice machine that will listen to you and like arbitrarily take notes or do whatever. And a Nintendo a video game console. And right like and doesn't it make sense to lock in the global gains by distributing the centers of excellence globally for a little bit here? And I think if you take a step back and you don't view it from a nationalist perspective, but you view it from a global economic perspective, that makes a lot of sense. And even then, we're not like I'll tell you right now, are we as good as the Chinese drone parts companies? Not yet. We're getting there. We're getting there quickly because we keep shipping more and more and more. And that hardware cycle is virtuous cycle, right? And is that good for the world to have distributed production of these things? Absolutely. Because what if something what if a hurricane nails Shenzhen and all that shut down, but farmers rely on drones to do this or relying on drone delivery or whatever else? Like from a global redundancy of the gains perspective and from a national sovereignty perspective like I'm not ignoring that but I'm saying even if you're somebody who totally disregards that it's still a globally economic good to spend this moment and create redundancy.
From a sovereignty perspective obviously you want to have critical technologies that are necessary for the defense of your society and of what you hold as valuable for your civil liberty you kind of want the means to be able to do that. So I think from that perspective we see it and but that's the idea that's not one you have to explain to people because we're associated with defense tech things and people are like yeah drones and they see the conflict but I think people overlook the fact that we're locking in these global gains.
You know, it's so funny that you say that cuz I've had people make comments about drones and, you know, I'm just all about war and this is like the farthest thing from my mind is war. I wish we had no war. I wish we had world peace. I mean, I really truly do wish that, but it's not reality. And the reality is we do need to make sure that we can protect this country and we need to make sure that we can bring our guys home safely, right? I mean those are things that I think are important and so to like you said I mean a butter knife at a table could be used to spread butter on your roll. It can also be used to stab somebody in the heart. It's a tool. It's I mean that's just the reality of it. Should we stop making tools because they can be used for good or bad?
Kind of don't think about it.
I personally don't think we should and you know I sounds like both of those are bad though. My cholesterol is not great. So butter on a roll or being stabbed sound like they're so I don't know that that's the best analogy given what I understand about my medical conditions but I think people might understand your point.
No, I speaking of defending infrastructure like really excited a lot of our stuff goes into counter drone you know we also did the investment in Power US and a lot of their Guardian stuff is being used to defend oil infrastructure and other things from non-state actors like the Houthis right not even Iran proper but a lot of the non-state actors who want to disrupt supply chains and you know like interceptor drones are a great way to prevent that and So, you know, it's really cool to work on the defensive side of stuff and, you know, the same application, the same parts, but configured to do things like stop oil pipelines from getting shut down, you know, like all that. It's a that was a segment we didn't really expect. Like, I didn't see that one coming and then once it showed up, it was, you know, we're really happy that we can help facilitate the supply chain for that,
Right? I think the physical AI thing is it's one of those things where you can if we start to apply it right we can then start taking like the maybe the life expectancy of a part and how long you know before it needs to be replaced. There's all kinds of things that we can do good with physical AI that put you in such a sweetheart spot. I can't even accurately probably describe it.
Yeah. Can we just say robots? Why do we got to use new words? Like and yes, motors and batteries and controllers and cameras, they don't just go into drones. They go into robots. And if you look, the FCC recently banned foreign robots, including things like robot vacuums.
So what does that mean exactly?
Well, it means if people don't get waivers, they're going to have to make robots in the US. This is the same playbook they used for drones. So now it's being used for things like robot vacuums and humanoids and internet routers because I think we've discovered and you can go back to the Israeli pager incident or more recently the Kraken reporting sending IP addresses back to China thing. I think what we have realized is that if there's any gap in the supply chain, the end product is not something you can have assurance over in the same way where even recently Trump was encouraging the Ukrainians not to blow up diesel plants because refining is limited to a few places in the world and so if they blow up diesel plants there's going to be less diesel because that part of the whole supply chain is a choke point. And so I think what the US is trying to do is really get ahead of humanoids and other emerging robots as we look at physical AI and saying, 'Hey guys, just make these in the US so that we don't fall as far behind or lose all the technology like we did with drones.' I think they can look at drones now and be like, 'Oh, we have so much catching up to do.' And we do. And they say, 'Hey, look, if we just do this up front, these markets will mature inside the US rather than have to we'll have to supplant an entire category of everything and you know, so I think it's a little bit forward-looking, but again, a lot of the parts used in drones are used in other robots. So, as we build this infrastructure for the drone market, this infrastructure, this manufacturing infrastructure and this component infrastructure that we're building will also apply to robotics. So if we distilled this into a business talk, then our TAM got a lot bigger.
A lot bigger. I think there were 8 million robot vacuums sold in the US last year. You know, they all have these little wheels that spin and those are driven by motors and they have batteries and they have a controller that detects things around them. Kind of like if you took a drone and just bent it on its side and put wheels on it.
A lot of the same parts. So I was sitting here being like, 'Well, I guess I don't have to do anything for TAM expansion except for not screw up.' And you know and it's because and I think in a very thoughtful way by strong bipartisan support the government recognizes how important this holistic agentic physical AI robotics layer is going to be to the future of the country and how fragile it is if somebody can suddenly take control of all these physical things. Like it's one thing to have your computer hacked and to even worse for like deep fakes to be out there whether they're videos of you saying nice things about me Tim or whatever, you know.
But I think it's a whole different concern once adversaries can take over parts of the physical world.
Yeah, it's always been a curious thing to me as to why my vacuum or my toaster has to connect to Bluetooth.
Obviously, your toaster wants to know when you wake up so it can start the coffee, preheat the coils. So, boom, perfect toast. You're set.
Like having everything interconnected is amazing.
Yeah. Well, why does the vacuum need to be I mean it's like what's the deal there, right? I mean, do you want to buzz around when you're on an interview with me or do you want to wait to go for a coffee and have it clean the house? And do you really want to have to tell it what to do or do you want to test the thing?
Well, I guess it depends on where it's made, right? I mean, but even people will then have issues. I mean, look at what's going on with Flock or whatever that,
Yeah. There, you know, there's a lot of concerns there. Privacy issues, stuff like that, right?
Yeah. I mean, people I think Well, that's a whole endeavor. I think people like the perception of privacy.
Yes. And I mean there is no privacy. We give it up for free to our TVs, to our smart speakers, to our computers, to our email, to our phones. And I think there's a lot to be said about just accepting that we're in post-privacy society and to just make an argument that you deserve anonymity rather than privacy. But I think you know I think it's an example of a place where people don't all of this stuff is really hard to understand. You know I think the world the evolving nature of the world and technology is bigger than you can hold in your head anymore. You know for me or for anyone like if I went back to 1950 I can probably understand or work on almost anything still and the world's accelerated so much and become so different so quickly that I think it represents a place where people have this absolute rage that authority which is the police might be misusing technology in a violation of privacy and it makes it a really easy target and it's easy to get people mad at things. I mean it's a common political approach these days. But you know I think it doesn't actually address the underlying either negative social negative aspects or sort of what we've actually accepted and given up.
Absolutely. Absolutely. I mean a lot of people have given up a lot since Facebook, Instagram, TikTok. I don't think people recognize how much they've given up with their private life, right? And they're okay with that.
It's awesome. I mean, like look, if I go ask anybody, hey, would you want to be a king or a nobility in 1890? Of course, I'd have a private chef and a da da and a da da. Yes, we've given information to the machine and we functionally have those things. Like I rate on Yelp, it tells me what restaurants I should go to. It tells me what I should order on DoorDash. I can go get a training plan that it'll follow it. I can use Duolingo or other apps to learn a language. It'll annoy me. Like we have the amenities of the most senior nobility of 150 years ago. But do you think that their servants didn't know about their lives? They gave up privacy for that benefit. We're just making the same exchange. So anybody who's like, I wish I was a prince or a princess, you are. You are already.
But yeah, you can't have that. And privacy because they sure did. Like somebody or something has to know you to behave as if they know you.
Fine. It's a great trade. I'm here for it.
You have such a refreshing look at life, view of life. I think it's so important because there's so many positive things happening. I see so many positive things happening in this country right now that I'm so excited about because I just see a very good future. I see us solving a lot of problems that big problems, not little problems, big ones, right? I
We just got to pull our heads out of our butts and start looking at the world a little bit differently, right?
I disagree with that. Totally disagree with that. I think the righteous, critical, consumer-driven nature of what we do makes us go so much faster, right? Like instead of accepting something that's 95% good, people complain about the 5% that's bad. So it gets fixed faster. That's why scale gets winning.
I think people winning is great. Like if everybody was like all these things are perfect, how would we get better, right?
