#49 - Ian Laffey, Cofounder & CEO of Theseus - Relentless: Summary & Transcript

In short

Podcast Episode Notes: Relentless #49 - Ian Laffey, Cofounder & CEO of Theseus

Overview In this episode, Ian Laffey, the co-founder and CEO of Theseus, discusses the innovative work his company is doing with drones, especially in GPS-denied environments such as those found in Ukraine. He shares his journey from hackathon projects to serious applications in military settings, touching upon the challenges and realities of developing reliable technology amidst war.

Key Topics Discussed

Introduction to Theseus
Focus: Building a drone navigation system that operates without GPS.
Technology: Utilizing feature matching with existing maps (e.g., Google Maps) to enable drones to fly autonomously.

Hackathon Origin
Hackathon Journey: Ian and his co-founder Sasha participated in a hackathon that sparked the idea for their drone. Despite having software backgrounds without much hardware experience, they collaborated with others to develop an initial prototype.
Initial Challenges: The early prototype struggled to perform effectively in real-world conditions, prompting further exploration and iteration.

The Importance of GPS Denied Environments
Real-World Application: Ian highlighted the significance of their technology for military applications, particularly in conflict zones where GPS signals are jammed.
Collaboration with Ukrainian Forces: Their work gained interest from various military sources, emphasizing the urgent need for effective drone solutions.

Iteration Philosophy
Rapid Prototyping: Emphasis on quickly developing prototypes and iterating based on real-world testing and feedback.
Learning from Failure: Ian shared experiences where early flights did not meet expectations, which served as learning opportunities for future iterations.

Engineering Challenges
Moving from Prototype to Production: The discussion highlighted the transition from creating a functional prototype to developing a robust engineering solution that could be reliably deployed in combat scenarios.
Importance of Reliability: Ian stressed the need for high reliability in their systems, especially for military applications, where failure is not an option.

Building Trust with Users
User Engagement: Ian talked about the importance of understanding the end-user's needs and ensuring that the technology works well for them.
Communication and Transparency: Ian emphasized the need to manage expectations and maintain clear communication with military partners.

Personal Reflections on Risk and Responsibility
Experience in Ukraine: Ian shared his experiences from multiple trips to Ukraine, discussing the intense environment and the responsibility he feels to deliver results that can save lives.
Balancing Innovation and Safety: He explored the tension between the desire to push boundaries in technology while ensuring safety and reliability.

Future Vision
Scalability: Ian expressed excitement about scaling their technology from small-scale applications to widespread use in various drone platforms.
Expanding Capability: He discussed plans to enhance their capabilities and integrate with a variety of drone types, emphasizing the scalability of their solutions.

Key Takeaways

Iterative Development: Rapid prototyping and iteration are essential in developing effective drone technology, especially in dynamic environments.
Real-World Impact: The technology being developed is not just theoretical; it has the potential to save lives in critical situations.
User-Centric Approach: Engaging with end-users and understanding their challenges is crucial for developing relevant and effective solutions.
Importance of Communication: Clear communication with stakeholders, especially in high-stakes environments, builds trust and improves collaboration.

Conclusion Ian Laffey’s insights into the development of drone technology for GPS-denied environments provide a compelling look at the intersection of innovation, military needs, and the ethical responsibilities involved in such work. The episode underscores the importance of resilience, adaptation, and the drive to create solutions that truly make a difference in the real world.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.

Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Drones and then also missiles all have GPS. Anything that's been built for the military in the last 30, 40, 50 years does its navigation with GPS. I want to make a drone that flies without a GPS to use Google Maps to like fingerprint where you are. We built that, put a picture up on Twitter and then slept for like 18 hours. I like woke up, opened my phone and tons of people from Ukraine, tons of people like associated with the White House. Very quickly became clear we needed to do this. This is Ian Laffey and he is the co-founder and CEO of Theseus. and theseus is basically building a little box that sits on the underside of a drone that makes it so drones can fly into like gps denied environments and basically fly without gps um you want to just talk about what that is and why it matters yeah yeah for sure so yeah you want drones to do things autonomously you want them to do things without needing an operator to control them the whole time uh and in order to do that they need to know where they are because if I if I'm trying to go from A to B and I don't know where I am at A and I don't know where I am at you know A plus one then I can't get to B.

1:05I know that you like went through a hackathon and basically let's start off there. How did that hackathon come together? What did you build and like what came out of it? Yeah yeah for sure. So yeah Sasha my co-founder he he was at he was doing a PhD at Yale and one of the guys that put the hackathon on was also like they were buddies from school. And we did our undergrad together. Like we're very good friends. And so he, I think I'd seen it on Twitter, but I wasn't like planning on making the trip down to LA. And so he, yeah, he messaged me and he's like, Oh dude, do you want to go to this? Yes, definitely.

1:42Neither of us knew anything about hardware, really. We both have like software backgrounds. He's got like an AI background. And so we're like, we got to find somebody there that like knows their shit with hardware um and so we went down there we met carl he's he just knew a ton of stuff about drones um and he's like yeah let's i want to make a drone that flies without a gps uh to use google maps and uh to like fingerprint where you are like dude that's like fucking sick idea uh and so we yeah we built that like honestly the early version of it the prototype was like super like it just wouldn't work in any sort of real environment like it was like very much like Like, okay, you're in an urban environment.

2:23Like, you're matching features to satellite maps that are up to date. Like, it's basically playing on easy mode. But we made that. I've done a lot of hackathons. And, yeah, they always kind of like results. I don't do them anymore because anytime I do a hackathon, like, I get a new job. So, yeah. You like the one that you have. Yeah, yeah. It's all good now. Exactly. Exactly. So, yeah, so we did that. um uh like basically didn't we did sleep for a couple hours actually uh in the middle and then like went presented the thing um we didn't win the hackathon or anything like we talked to people and you're like oh okay cool um and then yeah i like put put a picture up on twitter and then slept for like 18 hours uh and i literally woke like when i went to sleep the tweet had like maybe like six likes and i'm like in an uber like to my air it's over it didn't work yeah no i'm just like no I did.

3:13I like I need to pass out. No, at the time I'm like, dude, like six likes banger. Like this is like sick. And so, um, yeah, I just go to sleep and I like woke up. The first thing I see is like a message from one of my friends from college. He's like, dude, like saw that tweet. It was pretty cool. I'm like, oh, okay, that's weird. Like this guy hasn't texted me in a while. Like cool. Um, and then, yeah, like open my phone and I saw that and I was like, oh shit. Um, that was my first reaction, like shit. Uh, and then pretty quickly it was like, yeah, we, um, We really got to do this. Like a lot of like very serious people are messaging us and like want this and like want it yesterday.

3:48I think I heard you say something along the lines of someone from Ukraine actually messaged you just after that video. Yeah. Oh, yeah. Yeah. 100%. So we had like tons of people from Ukraine, tons of people like SOCOM people, tons of people like associated with different government agencies like the White House and stuff. So, I mean, not like anybody in the White House, but like, oh, I have been talking to the White House, you know, you need to do this. So, yeah, it very quickly became clear that we needed to do this. Like, yeah. One thing I really liked was kind of the iteration philosophy where you're just thinking about going super, super hard and really, really fast.

4:25And you're not also trying to like vertically integrate everything. You're trying to find parts off the shelf and then put them together to make them work. um how did that coming out of the hackathon you realize the thing doesn't really work um how to go through that yeah break that into a couple parts the reason that we um knew that it didn't work is we were lucky enough to talk to people in ukraine during the hackathon and they'd been dealing with this problem for a while because their gps's are all getting jammed um and like you know just to dive in that for a second because it it's like such a everybody's heard the term at this point.

5:01It's kind of like buzzy, but it's like, you know, when you're over there and you're driving around, like you've got Google Maps, it's like, oh, I'm going to McDonald's. And then all of a sudden your phone says like you're in Iceland and you're like, oh shit, dude, like where, you know, where are we going? It's like, it's literally like in the car driving around like anywhere, especially as you get farther east, like you're just going to get your GPS knocked out like all day, every day. So we talked to these guys who had been experiencing this for a while just to be really clear too the reason they do this is like um drones and then also missiles like all have gps so like anything that's been built um for the military in the last 30 40 50 years like going back to tomahawk missile turcom does it's like a large part of its navigation with gps and especially the terminal phase of navigation with gps so that's like the difference between um i'm gonna blow up like you know this block and hit somewhere in the block with a payload and like I'm going to blow up like this building.

5:57I'm going to come through like the front door and like you're going to die. So that's that's why they have to jam it. And so it's like super easy to jam. Ukrainians figured it out. Like, I mean, nothing you figure out. You just like blast. It's known bands. You just blast a bunch of radio interference in there and no more positioning. So they had to figure out and they took one look at what we're doing. And they're like, guys, like we've tried this. This isn't going to work like, you know, it's like you're flying over like fields. So like, what are you gonna, how are you gonna match features over fields?

6:26Like, um, things, things explode, they change like rapidly. Um, there's snow, all these different things. So we, we definitely like pushed that a little bit just cause we wanted to like actually put something up in the air and like have, have some proof of concept, et cetera. But like from basically from day two, I'm just like, fuck, fuck, fuck, fuck, fuck, fuck, fuck. like what are we going to do like the way we're solving this isn't going to work um and so i basically just read like every single paper that i could get my hands on and we're lucky also like we've been continuously lucky but because um because we had like some some press and exposure and people thought what we're doing is interesting and also like yeah just kind of associated this problem with us we got to talk to a lot of like researchers that were working on the problem so like got to talk with people from like MIT, CMU and like bit by bit you like piece it together.

7:19But like, especially the first conversations you're talking to these guys and it's like, you know, this guy's been doing this for like five years. Like I did this at like a hackathon. Like, you know, like I'm just a little bit of tribal knowledge that isn't really there. Yeah. Yeah. A hundred percent. Um, and so, so we, we got through that pretty quickly, thankfully. Um, but, but yeah, then we, we kind of landed, uh, we, we did like our homework and, and landed on like a, uh, an approach after I would say a month and a half. I mean, actually there's a funny story there, which I don't, um, I don't think I've actually talked about, which is that, uh, so Sasha was doing a PhD.

7:56Um, he is French. So he had to get his like, Oh, one visa, get all that sorted out, um, and then drop out. And so he, he ended up moving to San Francisco on April 1st of last year. And he showed up like at my apartment. And we're like, okay, like, let's go. And we're sitting there and I'm like, dude, like, this, like, this isn't gonna work. There's nothing to go on. Like, we got to like, go read papers. And so like, that day that he was there, like, we found the thing. And we're like, okay, like, this is how we're gonna break it out. Like, you're gonna work on this, I'm gonna work on that. Go, go, go, go, go.

8:31And like, that has been the like, high level approach that we've stuck with since then. But it didn't come together until um i guess like six weeks after the hackathon yeah i think one of the early iterations was just strapping like an android phone too what was that yeah dude um so that was the first approach we took and there's a couple funny things like um yeah i didn't want to i didn't want to figure out how to run it on the phone or like do it on the edge so it's like okay we're gonna host this on AWS, where we've got these big GPUs, and we're just going to send like an image from the Android phone to AWS.

9:09It had this like crazy latency, obviously. So it took like a second to come back. And we didn't like when you're doing like, you know, GNC guidance, navigation control, you need to do a lot of you get these inputs, they have different timings, you have to filter them, you are constantly filtering between your different sources. And you hear like buzzwords around like sensor fusion. Like there's, there's a lot of math and like stuff that goes into this, which we have since learned and hired people, hired people that are really, really good at and ourselves gotten good at. Um, but to begin with, it was like, ah, fuck it.

9:44Like we're just going to like, we're going to send this to AWS to take what it gets back. We'll just like send that, um, over, over a serial connection into the drone for like two seconds at a time. And then when, when, when it updates, we'll send a new one. Um, and so we put that, I don't know the statute of limitations for like flying drones over college campuses but like i will just say this because i think it's funny it's like we put that up like over stanford is like they've got these like no drone signs so we're like okay we're gonna go like away from the no drone sign we're gonna go like a couple hundred feet away from the no that's legal yeah and we just fucking put it up in the air it was uh it was a holly bro x500 one-on-ones there it's like everybody's first drone um actually like our very first angel investor was we hadn't like figured out like I don't even know if we'd incorporated or we hadn't done all this stuff so we couldn't take his money but he's like I will Amazon you like a drone what do you need so we had this drone that he bought us we had the Android phone that that I bought from Cricket Mobile on Mission Street and we just like slapped it together put it up in the air it was so fucking shaky it was so scary but like that was like it was pretty quick and so that's what i was like happy about is like okay we are like you know we are actually flying this and like it didn't work but like okay look like we did that now let's go and make it better from there yeah i think i think one of the big things that people have kind of gotten wrong or maybe directions they've gone down uh with you know autonomous cars and such is basically slapping a bunch of very expensive sensors onto you know a car you know hundred thousand dollars with the sensors onto a car um and i think you guys are taking much more of the like tesla approach where you're saying like okay we use visual line of sight for you know humans um and we're just going to strap a bunch of cameras that you can buy anywhere um and then use basically come make up for that with software um and i i like one of the things that you said where you basically said we're going to take in bad not necessarily amazing data but we'll just make up for that on the software side.

11:44How do you kind of like come up with that? To me, it's like significantly harder to get good data from the real world than it is to just make bad data better. I mean, that's that's what do I mean by that is like to solve these data problems, get higher quality sensors, get LIDAR, et cetera, et cetera. Like there's very real like engineering there. And it's kind of like a meme, like software engineering isn't real engineering. But like it is actually true like as a software engineer starting to touch like real engineering you see very quickly that like um with software you're you're kind of in like la la land where you can always like spin up some new thing it's like you have some trade-off around like compute and cost but like you're literally like in not not the cloud like as it's used generally but like you're in this this abstract space where you can do whatever you want whereas like the second you have to do like real engineering it's just like you have all these physical hard physical constraints um I think the other thing with like self-driving too is like self-driving is a much harder problem in almost every way than like than drones because you have this like regulatory compliance.

12:55You have these safety concerns like if I always thought this was like a beautiful problem to attack like and to start a startup in because the drones fall out of the sky all the time. Like, you know, like it was like the default. and falling out like an isr drone which is like designed to be used and reused and reused again um like the average isr drone in ukraine like i'll fuzz the numbers here i think they're like publicly known though it's like say 25 to 30 flights like per drone so it's it's just not like if your waymo exploded or got shot out of the sky after like 25 flights or 25 like rides like you'd probably have to think about it differently um you see the quality cash burning in flames and You're like, we might not this might not be the right business model.

13:42Yeah, I mean, it's that. And it's also it's just like it is literally it is war, man. And it's like the worst thing ever. But it's also like an extremely, I want to say, like interesting engineering environment because you don't have regulations. You don't have yeah, you don't have the FCC saying you can't do this. And you don't have, you know, the DOT telling you that you can't fly here. and you do have to like coordinate like it's it's not really the wild west especially these days like out there but it is like a lot of the rules that we have around technology are pretty arbitrary and I mean I saw even like yeah I saw some somebody recently I think it was Isaiah talking about like regulatory hurdles like in the US there's a lot of regulatory hurdles like the second you go to war, like all regulatory hurdles, you have these these organizational, logistical, bureaucratic issues.

14:39But then like when you're actually out on the front, like there are not really like, did you get the job done or not? Yeah. And and and, and it's almost like, I think we've been learning like what those because because it's there are really big bureaucracies at play and you do have to learn those processes. But it's it's a very different motion when you have like a commanding officer that you're responsible to and you have guys underneath you that you're in charge of. And you you need to achieve a mission versus just like, oh, this doesn't seem safe. I'm some official and I'm going to kind of deem this unsafe.

15:16So I think it's just like a very different environment to develop. It's like the difference between like optimizing for like in users and their ability to execute their missions versus just some guy that's like stamping stuff. Yeah, yeah, yeah. We've, yeah, we've spent like, pretty much all of our time. And I think that this is kind of not what most, I think it's what a lot of the new defense startups are doing. Like I heard in your interview with Olaf, he called this out. I know, like, Nero's, they do this as well, where they, they do this even more than us, where they spend a ton of time like with end users, like really working with that.

15:48I think that comes very naturally to people that come from like an engineering tech startup background. But it is actually very new in the defense space because, yeah, it just, when you're procuring, and I think some of it does have to do with drones and the nature of drones. Like you have these small edge deployed systems that are used at like the squad level, They're used at the platoon level. And so you like if you have an F-35 like that, that's such a strategic asset that like the user of the F-35 is not the guy piloting it. Right. It's the general. It's the I don't even know at what level they control that, but probably like two, three star general that's saying like, you know, go go do this, put this here, do this thing.

16:35Whereas like with drones, everybody has a drone. It is like your you have a team of six dudes and these are your drones and you're in charge with using like in charge of using them. And so I think that that is like a lot closer to what you're used to seeing from startups where it's like, go work with the end user and the end user is the person that is using the thing. Do you have any other really good stories of kind of deciding to move much faster and just make something out of, you know, off the shelf parts? Yeah. What can I say there? I mean, I think we just have always felt it's actually gotten a bit better.

17:13the past six months, but the first like year, we just really felt, I mean, again, there's a war and like we're supporting people that are like really in almost every way, like kind of the underdogs there. And so it is, it just always felt like we needed to like, it's not even, it wasn't even some conscious decision of like, oh, I want to move fast because, you know, things compound over time. And if I do this in three months instead of six, then, you know, over the course of five years, I'll be so much farther ahead. Like it was not any sort of thought like that. It's just like, oh, like I have this guy on signal that's messaging me that, you know, I told I would be done with this two weeks ago and I'm still not done with it.

18:01And that feels really fucking bad. And, and I've come to learn that that's like a super common thing. And, and it kind of like sucks to have been on the side. I think what, what makes it better for me is like getting through that and you actually get out there with the guys and it does work. But like, I mean, this happens all the time with, with people in Ukraine is, is that they hear some, oh, there's this hot shit Western technology that's going to come in here. And then it just takes longer and longer and longer. And so I think for us, it was just a lot of like, yeah, I, I don't know. I think what has really sunk into me about working on this is that it's not like, there are no countries.

18:44There are countries. I believe in America. I believe in the West, but it's like people. You're not working with a country. You don't work with America. You don't work with Ukraine. You work with like six dudes. You work with like one guy in like an office somewhere that like has a family and like his, like people are dying and like their brothers are dying and shit and so I just think like when you have to um when you start interacting on that level it's like much less um yeah man it's just like yeah my friend like my friend is in a bad spot and like I I need to um there's a real fire underneath you for like making making thing work in like not only the time frame that you like said but maybe like a few days earlier and like not having that happen really, really sucks.

19:31Yeah. Yeah. I think it's just been, um, that more than anything. And, and to me then it's like, um, yeah, I always look at like, what is, um, you said this when we're talking earlier, but yeah, like what is the critical path to get to something? Like what is the, um, what are the things to de-risk and then what is the way to do that as quickly as possible? Um, and also like, is what we're doing actually working? Because I think so much of like, um, I think you were alluding to this as well as like so much of time is spent doing things that like if you reflect on it, you'd realize it wasn't going to work.

20:07And so you need to constantly like meta prompt yourself of like, is the thing I'm doing the right thing? Is it going to work? Have I, you know, OK, I've been working on this for three days. It feels really fucking good to be writing code and like see this boulder being pushed. But like, am I like, is this actually the boulder that I care about? and like you know when you were definitely pushing the boulders just is it gonna ever go downhill like doesn't matter yeah and so I think we um yeah just constantly like running that prompt and and um I think a lot of times like we did well at that and sometimes we didn't do well at that and I think like that is uh that's kind of the key thing is just if you're working on the right thing then uh you're gonna you're gonna iterate much faster you know you tried iterated whatever or you put the thing up in the air over the Stanford campus.

20:57Coming out of that, how do you kind of think sort of the critical path in your mind? Yeah. So the first time that we went out on base with SOCOM was, I think, July of last year, it was either June or July of last year. And we were able to do that because we were able to put our sensor onto the wing of a plane and record a data set and then localize over the data set. And then I showed that to the guys and it's like, oh, this was an experimental aircraft, by the way. So that's legal. And it actually was. No, like I'm not just saying that like it was an experimental aircraft. But yeah, so we did that and sent it to these guys.

21:38And then, yeah, once you have once you have the people that you can get check it against, then it's just like, OK, what do they need next? And so we were able it's like, OK, we're going to we're going to go out there and we need to fly this autonomously. and so this needs to go on the bottom of the drum that they have it needs to fly autonomously um and then yeah i mean like i honestly okay here's to really answer your question how do you think through the critical path um i don't think through it at all i do not it it is um i just have i think a pretty good gut sense of what good looks like and i try to calibrate how i'm feeling to results and then reflect on things that I've done in the past and then update my feelings based on that.

22:25And so I really just like think about it. It's like, oh, do I feel good? Do I feel bad, et cetera. And like recruitment style where it's like, I know what I like and I know what I don't like. Oh my God. So I was, yes. I mean, we literally, um, so much of the shit that we did, like we just, we just had to learn ourselves and like a lot of it, one of the, the biggest accelerant that we've had i mean obviously i think so for any company it's just been like hiring people bringing people on that are really good at stuff and can help you with those things but uh to begin with like yeah i mean just figuring out like a lot of stuff from from first principles and with not enough like background research just like literally just like you're just trying it yeah yeah i um yeah i'd give an example on this i think it's like maybe a little too nerdy, but yeah, like in robotics, you have like, you have coordinate transformations.

23:20So like you need to go from like, you're oriented like one way to like another way. And so there's like, and I know linear algebra pretty well, but like, you know, you just, you multiply a matrix there and it does the rotation. And this is like, you can study these coordinate transformations. Like they're not hard. Like it literally takes like a day to learn. But like this wasn't something that like either myself or Sasha had a background in. And so we literally, I'm just like, dude, like, I think we need to multiply by like, I think it needs to be like negative X or like swap the X to Y. And you're just like, and you look back at like that first time, it's like, dude, holy shit.

23:56Like there's just, we didn't have to do any of this. But I think that that helps a lot to kind of think things through that way, because some things are really silly. Everything is really simple when you know how to do it. And most things somebody knows how to do and you can just like read a book and do it. But then some things are like actually hard and nobody's done them before. And so I think like that that's been a funny thing is looking back and I was like, oh, that thing that was simple and but took a week was actually novel. And that thing that was simple and took a week was actually kind of silly.

24:32I think a lot of that. Yeah, I know. I know you mentioned and Christian's podcast, I think you said something along the lines of you really try to like go find the source of truth or like someone can is, you know, a few steps ahead. Like I think Mr. Beast even talks about this where he's basically like, go find not the consultant, like not a McKinsey consultant, but someone that just really knows, you know, if you want to go make the biggest pizza in the world, find the person that has made the biggest pizza and then figure out like, you know, how to do that. So how have you kind of used like finding those people that are like a few steps ahead of you?

25:07How have I done it to your advantage? Yeah, I yeah, just I just like I like talking to I like talking to smart people and I like asking questions. And I think most especially like successful and people that have done things are very happy to talk about it and share it if you can ask them the right questions or get them going about the thing that you're interested in. And so, yeah, I think we just always run this process or I specifically always run this process of, yeah, we're, it's funny that you mentioned that. Like I don't even consciously think about it that much, but it's like, yeah, we're, we're doing this, we're doing X.

25:52Who do I know that's the best at X. And it's, you know, there's, I have a large like contact list in my phone. And there's people that I go to for like a lot of things and like general things. And I have a lot of friends that are founders. And if they're, you know, series A, series B founders, and I'm like, it's just some like general company startup stuff. It's like, okay, I want to talk to this person about this and get advice. Or, yeah, like in industry and with technical problems, like people, good people know good people. And so I'm just, you know, I will text so and so, you know, Johnny or whoever and say, Hey, who do you know that's good at this?

26:31Or like, these are the things I'm thinking. And I think the last thing I'd say there is, um, I just, I really try not to have an ego about anything. And I really try to like ask the stupidest questions that I think of. And it's like, um, sometimes very occasionally you meet people where you're like, they're really scary. And it's like, you're you know, you're like, fuck, I don't want to like ask this and you kind of miss those opportunities. But like, almost always, I'm just trying to like, ask, is there aspects to it where you're basically trying, you're not worried about the ego side, and you're like, Okay, I may look like an idiot.

27:06But it's possible that it's actually the stupidest questions actually probably are pretty smart. In most cases, you're thoughtful about it. Almost almost always the stupid questions are the good questions and and almost always the ones just when you feel nervous or you're scared or you're out there like and you just feel kind of like an idiot then then you're getting somewhere and also I think the thing that helps is just like when you have um when you have a history of past execution and you have actually done something then you don't need to feel bad asking people questions um and that's something that just like really does compound over time it's like it's like building it's like a building of proof of work so that when you go ask that thing the person takes it seriously well and it's like, dude, it's I mean, I kind of started there because I don't actually know what I should say.

27:50But it's like, yeah, like this shit is flying in Ukraine. You know, like it does work. Like it, I know that it works. And I know that I've done the thing. And I can tell you all about it. And so, you know, I may have some question about some new thing that I want to do or some something else. But it's like, I think that gives a lot of confidence in terms of, yeah, I'm gonna sound like an idiot, I'm going to talk to you about this thing that I don't know anything about. But honestly, like, and it is kind of ironic to say this on a podcast where I'm just talking about things that I do know. But like, I would say 90 plus percent of the time, it's like, why am I saying something that I already know?

28:28You know, like, you need to spend, yeah, maybe 10-15 % of your time, like, summarizing to other people. And like, there's been something that I've been learning as we've like, been growing a team is like, these guys do need context, I need to do a good job of like providing them context and and and giving them something that they can rally behind and understand but other than that it's like I don't know I think everybody knows that person that like comes to mind where it's like they're just telling you a bunch of shit that they know and it's like what like if you're not like what's the point of that like you should I enter I try to enter most conversations um again this is ironic on a podcast but like I generally you know where I'm talking about my own stuff but it's like I generally enter most conversations conversations like what can I learn from this person and whoever's doing I can think of like I got a buddy Peter very very smart guy and like if you let him like you'll talk to him for 30 minutes and he won't say anything about what he's doing he'll just ask you questions and he'll like because he's just like he's so smart he's just like training his super curious and and and just like if you're the one asking questions like updating your mental models like you are just like gaining a lot from that conversation.

29:39And so, you know, I could sit there and just say the same five things that I've said before, but it's like, you know what? Yeah, like maybe you get somebody thinks marginally. I just I don't know. Maybe it's a way to get people to. If you care what what people are thinking about you or something, but I don't think it's like the best way to build like your own like model of understanding of the world. I think you just need to try to learn a lot more than you um than you share and then also try to share when you're like going through the first few iterations of your making now uh what did that look like it looked like collecting a lot of data and then well it looked like reading a lot of papers collecting a lot of data, training some models on that data, iterating on those data sets, then from there testing in the loop and then iterating on the real world results.

Read the full transcript

30:54Because like our product does not exist in a vacuum. It's like an input into like a larger system. It's very important that we understand the larger system very, very well, and then test the entire system. And so then it moved from this kind of like Eugene Stoner in the garage type thing to like, you're actually taking it out and just testing, testing, testing, testing. And the first hire that we made was like literally somebody to run like the testing arm of the company. And that's like been the most, like that was that was to that point the most impactful thing that we'd done um it's like get somebody to come on and help us test this how'd you find that person he messaged me on twitter yeah yeah we put up job postings and uh yeah Grayson uh reached out and she's like yeah man like I really like what you guys do are doing like I just want to work really hard like I don't know if I have the most experience in this but like I just really care about what you're doing and I'm gonna work my ass off um and that he did and um you know now he's yeah he's built a bunch of drones taught a bunch of people to build the drones now he doesn't even build the drones like runs runs all of that has been to ukraine like has done the integrations over there like yeah it's um but yeah just wax on wax off on the testing thing.

32:21And then also like, up until, yeah, up until like May, things would still not, there were things that worked and things that didn't work. And so I think there's an impulse to do more of the things that work. And it's like, no, we just need to do the thing that didn't work over and over again until it does work and so you know instead of yeah i i think that was like that was the big thing is just testing and like okay edge case boom fix it okay this thing okay fix it um yeah i know that you've like you have gone to ukraine multiple times i think you just came back from friday um and so how have you tried to kind of build that that trust with those people on the other side um yeah up to date yeah you gotta you gotta over communicate and you gotta under promise and you gotta over deliver.

33:13And I think that was like something we got we got hip on pretty quick is just like, yeah, you don't need to promise anybody the world like you just need to go and deliver on the things that you said you were going to deliver on. And so yeah, like, sure, first first month, first month and a half, we were naive there. But then from there is like, okay, like, this is exactly the state of the system. Like, okay, this, this, that is working. Like, we're going to come out at this time. These are the things that are ready. Please have this ready. Like, we're going to do this. And just being like, you just really need to like own the process.

33:50And you can get away with it. Like, and I think it's the difference between like an engineer and like a business owner too. It's like, I can, I can show up to your spot and just let you tell me what to do. And like, we did a lot of that when we first started. But then it's like, okay, actually, I've done this a dozen times before with a dozen companies that look just like yours, or with a dozen units that look just like yours. And I know what your drone looks like, I can guess every single part in it. And I can just build that. Like I have, I probably have a drone here that looks just like your drone, I will bring it to you.

34:24And we will work through this whole process together. And I will just like explain to you how long that takes and like what the risks are. And I think a lot of it too is like, again, just complicated stuff and like non-trivial environments and you can get into sketchy situations. And so I think you always need to be, um, and there are often other adults in the room, but you always need to be the adult in the room saying like, oh, I know that you may think that this is like how this works or that this is a good idea, but like, actually we need to we need to really do that um and and to the point about like showing up face to face is like i i think that that is uh i only i i think this because i've heard it enough times that i just have to believe that it's true is that um like as founders of the company people do like really rate when you come out and spend time with him and it's like oh okay i'm not sending like one of the guys out and then i'm like back here doing this this and that um and and at some point you have to do that because you're you're working with so many people that like I can't be in four places at once but especially to begin with just like a showing that you do care b yeah just I think being really earnest like every I think people have very good bullshit detectors and some people are very vocal about them and some people you know are maybe trusting but but it's just like yeah just don't don't uh bullshit people and show up and tell them what you're going to do and then do it and then when you build it's also a very very small world so like we're moving into like um you know larger oems and big army but like to start we were really focused on socom which is like a very small group of dudes like first of all like special forces like across the world like a lot of them know each other because as as part of special forces they train each other.

36:15So like you'll have Green Berets or whoever going over to the UK or Ukraine before the war or Poland or any of these places or South America or Africa and training soft people over there. And so it's just a very small world. And so if you like, if you do something with one person, like somebody else going to hear about it, like obviously, and you're only like max two hops away from somebody. Like you're in all the same group chats. Like I receive, there's, there's pictures that I'm not allowed to like send people. Cause it's like, Oh, this is sensitive. I'm not going to send this. And then I get that from, you know, and it's like, Oh, I didn't realize like you guys knew each other.

36:55Okay, cool. Like, let's go. And so I think just showing up and communicating well and like, yeah, understanding that it is like a small world and like reputation does matter. And like, we have, we have made mistakes. I'm sure we will make mistakes in the future. Like we, we have crashed drones. We have done things like less good than we should, but also we've done a lot of good things and we've tried to like learn from the mistakes that we've had and also just be very communicative with folks. And so I, I think honestly, that's one of the things that I've tried to like, um, really, really, really, really hammer is just like, yeah, under promise over deliver and over communicate.

37:36Um, just tell people why you're doing shit too like everybody's very aligned you know there's an interesting story of i forgot which well company this was but there was this guy that ran a well company where every single time someone gave like a here's what we want to do or you know here's what i want you to do um you had to give the why behind the action um and you just basically could not if you gave no why behind the action you just be fired immediately um because like people have to understand why they're doing a thing you know do you on the like earnestness side do you have any interesting stories from just being very showing up I'm trying to think what what I can say and then also yeah cuz it's it's always funny like what there's stuff that you can't talk to anybody about then there's stuff that you can talk to your friends about there's stuff that you can talk about and there's stuff that you can talk about on the internet so I'm trying to think about like what what is the internet okay yeah yeah as far as as far as earnestness um no like honestly top of mind i i think a lot of it is when you get into more sensitive situations and you just um i i don't think anything anything comes off top i think it's it's also like one of those things that's um just get meta here for two seconds like it's it's not it's you're not actually being you're not doing anything to be earnest you're just not doing something which is like bullshitting or like blowing something out of like it's it's just um yeah i think just trying to understand the truth and then and then also and this this applies to like asking questions this applies to working with customers i think it applies to everything in life is like phrasing phrasing is everything so like this is how i am thinking about this how are you thinking about this does this make sense to you or like because of xyz like you said about the y like because of these things i think this like check my assumptions okay assumptions make sense let's let's go um yeah so nothing like i wouldn't say any like funny concrete examples like come up, but, but it is just, I mean, here, here's what I can say is like, yeah, you're, you're selling to, um, we're selling to the DOD.

40:02We're talking to people there and just being like really honest of like, Hey, here's the results. And like, I'm showing you these graphs and I'm showing you these flights from Ukraine. This is really great. This is how many we've done. This is exactly how many kilometers we've done. Like we are not like even look at Neros. Like we're not Neros. Okay. We haven't shipped out like 2000 drones a month yet. We have done these things. And this is the list of things that we've, we've actually really executed on. And like, this is how it's increasing month over month. And I think people appreciate that a lot more than like, when you make it out, like you're, you know, I always joke like, yeah, like Johnny Ukraine.

40:43And then it's like, Oh, how much have you actually flown there? And so it's like, if you're talking about these things, just just make sure that you're you're actually doing the things basically. Yeah, I think I think Zipline had flown something on the order of like 50 million miles before the US even allowed them to like before they made their first flight in the US. Some crazy number like that. It's interesting on people who really want to move fast. They always kind of think they're moving slow just inside. And I think that's kind of what pushes them to move faster for you yourself. what have been the biggest moments where the biggest things that you do um to kind of try to move things make make timelines shriek you know move things forward that maybe shouldn't have happened in that time frame but you're just trying to push on that gas yeah well i just ask myself what is fucking me the most and then i try to get rid of that thing um and give any examples of that yes and i'm gonna give a bit of context before i dive into this it's just like we had a somewhat working product before May of this year.

41:45But what is really clear to anybody that is like in the space, like a somewhat working navigation product is not a working navigation product. Like you, you need to know where you are all the time. You need to not crash the drone ever, right? Like that's, if, if you're, you, you need, you know, like AWS, like seven nines of reliability, whatever, nine nines of reliability, like you need to, you need to view what you're doing like that. You cannot be 80, 90%. Oh, it worked all these times. It didn't work. So when I say like, oh, it didn't, it didn't work until May of this year. It's not like, oh, we'd never flown it.

42:21Oh, we hadn't flown it in this place. We hadn't done this, that it's like, it was not like, you literally take this thing, you stick it on the bottom of your drone and it just like, it just goes and flies. And like, if, if it's not working, like you, like something is misconfigured you tweak that and it just works so like until like me this year are you saying basically you didn't have enough field tests where you were comfortable kind of saying like no we we know this works i think it's just more like we didn't have that reliability to the point like like again it's like if it works on nine flights out of ten like that sucks it needs to work on ten flights out of ten it maybe it needs to work on 998 flights out of a thousand um and you need to be like pushing towards that.

43:09And if you're going out and like, if we're doing tests here and we're crashing, like we had a crash recently and it's like, oh, okay. Like, yeah. Like why? And it's not due to the system. It's due to like, oh, okay. We didn't, we didn't prep properly with the battery, but like, that's not good. We definitely don't want to have like an issue with our system. Like it should just not be happening here because you're going to take it into a much more complicated environment, not just like, oh, there's jammers, like jammers don't affect this. It's a camera. But it is like you are you are taking fire.

43:43There's drones whizzing over your head. You're pulling this out. You haven't slept in three days and it needs to just work. Right. Like there can't be any like fiddly bullshit there. Like it just needs to be like a fucking like a rifle. Like it just needs to work. But so to get to get to that point. Yeah, like May of this year, like I was just looking at the product and where it stood and all these different efforts that we were pushing, but it's like product just needs to be better. It just needs like, there's this, there's this big component in it, this big sub component that's taken us a long time and we need to finish this and we need to iterate on that.

44:23And so, yeah, we just went to Florida for a month. I told everybody, I don't even know how much notice I gave, probably not as much as I should have. It was like, I don't think it was a week. It was like, guys, like we are all going to go to Florida. Like we got a buddy who's got a ranch down there. He's got a drone company. We're going to set up there and we're going to fly every day. And we went to the shack that didn't have air conditioning. It was a hundred degrees and we had Starlink and we just sat in there. And it was like, um, you know, a couple of people would like work out of Airbnb some days, but it's like, I really want everybody sitting here in the shack writing code in the shack because then the second you finish something you you don't have to drive an hour like we're in downtown san francisco and like honestly that it kind of sucks right because you can't go do stuff yeah you can't go fly like it's like i that's what i realized like it's uh testing we have testing days and you i don't want testing days i just want to like test so if i have to sit there for two days and like work non-stop that's great but then also like i can fly whenever.

45:29And so that was like the biggest thing is like super tight iteration loops. Exactly. And, um, and whenever we've been able to move quickly, it has been just like, what is the, what is the thing that's fucking me most? What is the longest part of the iteration loop? And like, how do I cut that down? And, and frequently it is like one thing, you know, that's like 10 times slower. It's like, we got to drive out, we got to set up all this shit. um and and so yeah that that i think was like the most clear-cut example and like first honestly too like i don't i don't i talked to sasha about this i i don't even think i i said anything to the team but like i was pretty i was pretty worried like we'd you know like a lot of things have been changing like we'd we'd had uh like this big opportunity and like didn't go great and i'm like we really need like we really need shit to work like this is like it it's been long enough Like we need shit to work.

46:22And you get everybody out there and you're staying in this Airbnb or in rural Florida. Like it's not it's not San Francisco. It's not nice. You know, it's not, you know, beautiful. A shack is a nice term for it. Yeah, you know, it's a it's a I mean, the place is great. Thank you, Roger, for letting us out there. That the drone, the drone zone itself was great. But like that the Airbnb we're staying in had like fucking bugs and stuff. Like I'm I'm from Missouri. I don't mind bugs. But guys are like, oh, dude, there's bugs everywhere. I'm like, dude, it's Florida. But, you know, I was like I was like, oh, fuck.

46:49Like this needs to work. Like we're here for a month. Like we, um, yeah. And I was honestly pretty worried. And then like a week goes by, you get everything set up. First week is just like, okay, we're able to fly more like this kind of cool. What's happening. Second week, you start like ironing out these issues that we just had kept coming to. And it's just much, much, much quicker. And then by the third week, it's just like, oh, okay. Like, yeah, we're just flying through this. I'm like, we can fix all this shit and it, and it just works great. And so, um, I think, yeah, I think that that was like really one of, yeah, anytime we've been able to do something like that, it's been huge.

47:28After Socom was like the next customer and how did you kind of build that relationship? Yeah, we've got so many NDAs that it's like super unexciting to talk about. But one of the big focuses for us has just been like people that make drones. So you can sell to the end user who is using the drone or you can sell to the army. But really the way to get on those contracts is to go to somebody that's going after a big contract. And pretty much always they come to us and they're like, hey, we need this capability. And then we work with them. And so that has been that. And then and then some other stuff over in Ukraine.

48:15And again, it's like you can't. Like I can talk about this offline, but like, you know, for the world, it's like, yeah, like people want it and they buy more and then they buy even more. And we get to work through like the fun of integrating that and doing more. But yeah, that's kind of a kind of a boring answer. but like yeah NDAs and like yeah we're I you know I already spew enough stuff on the internet that's probably not one to talk about too much is there any other times where you were really just like waking up in the morning thinking to yourself man something is fucking me something is not working oh yeah all the time all the time and and another a lot of my quotes um are profane and use the word fuck but like i say these things to myself all the time but it's like yeah the thing that's that's fucking you the most is the thing that you don't even know is fucking you so like that if if if you're honestly if you're waking up like sweating and like have terror and like that is a blessing because unknown unknowns it's a yeah and it's it's a startup man like if you know everything like why why is this not you know a multi-billion dollar company yet like you know what's what's going on.

49:33Clearly you need to figure stuff out. And so honestly, that's the loop that I try to run is like, if I'm sitting there and I don't have that thing, like that's actually the scariest moment because then, then you need to, you need to know what the next, you know, one, two, three, four, five steps are. Um, but yeah, it, it's kind of just like how I view the world. Um, and I, I think like what I've been trying to shift as well as like it's really great and awesome like engineering wise to like have that as like a milestone. But then you reach a point where and with us, just because the technology is more difficult to develop than some, less difficult to develop than others.

50:19It took us a while where the technology was like the key risk. but you reach a point where the technology has been de-risked sufficiently to you have to zoom out and say okay like I trust the product team I trust the engineering team to execute on this and here are the goals and we're going to do this and map this out and I know that that will happen and that we're capable of doing that we have the people to do it and and you need to start looking at like basically engineering the company around this thing that you've built and I think I think you know, really good people kind of map that out as they're going for for us and me, I really wanted to just really just wanted to have the thing fucking work before like going and moving on and spending a bunch of time in these rooms with people.

51:10But but yeah, now I think it's a lot more of like, okay, you have this box, it does a thing, like, where do you go from there? Like, how how is this you know 100 million dollars how is this a billion dollars like how do you sell how do i sell this as a as a seven figure contract like this month you know like what what are the things um that we can do there because yeah it's like you know if a shark's not swimming it's dying if a company is not growing it's dying so um it's really not easy to sell things to the government. It's like a highly technical process. And so like, we have a lot of really awesome stuff in the works.

51:49But a lot of that is like, again, yeah, de-risking like it is just, it's not. I mean, it's certainly not like a stripe, you know, self-serve checkout thing. It's not exactly like three clicks. Yeah. I mean, and we have a stripe link, like you can buy this on a stripe link, like email me and I'll sell it to you on a stripe link that works but like to go from that stripe link like here's two of these to like okay like you know we we want to sell thousands of these tens of thousands of these hundreds of thousands of these and and millions of these how do we engineer that because you're you're not going to stripe link more than 50 right like it's just not that's not no one's no is buying like you know 100 000 every stripe yeah yeah no no and and um yeah so that's been that's been a lot of my thought process recently has there been any moments where like only in hindsight did you realize that you were completely like fucking up or you know there was something that you should have been you know waking up in the middle of the night over you just didn't realize it was an issue at the time I think not all of them are things that like are like can can necessarily be talked about but yeah yeah yeah I mean it's a great question if I if I can't answer it as it means it's like a really good question right because like you can't um in in and I think this is different in other businesses and I'm sure it'll be different like when you grow to like I can say the thing about about Stanford now because that happened long enough where it's like okay fuck like if they really come yes police are gonna come knocking it's like not really okay sorry guys like we we haven't done that in a long time but But I think, yeah, I don't know.

53:32I remember like being a kid and like, you know, when I was like eight, nine years old, like going down on my like buddy's older brother's Xbox and like playing MW2. And it's like, oh, M rated game. Can't do that. Five years later, it's like, oh, yeah, mom, like when I was eight years old, I was doing this. So I think it's like right amount of time to start sharing like the real, real fuck ups. Yeah. Yeah. Okay, outside of the Florida experience, have you ever done any other sprints where you're really, really focused on just doing field tests and getting those iteration loops? Yeah, we went out, both Sasha and myself, which was like, so this was in July.

54:15the team like in order to get uh the engineering for this uh sensor which has like this pick rail and it's it's all in one um and to get a large number of those out to a customer like the whole team here was like super focused on that so it's like okay like we got the requirements on this like you guys like i trust you like go go go go go so they're like super focused on that and sasha and i are just like man like we got to get back to ukraine like we got to go over there we got to bring like drones with us and like come and just and do that and so we went out um we went out like a bit near denipro um and we were there working with uh there there's tons of different groups out there but working with uh one of them out there and um yeah just like working through that whole process of of understanding like what they needed and what they were trying to do and use it for and then demonstrating the capability.

55:16I mean, even just getting drones. It's funny because it's very complicated. A drone is a very complicated system. The box we make is very complicated algorithmically. It's very complicated hardware. It's very complicated in general. And then also, literally, if you wanted to ship toothpaste from here to the front line, that would take two weeks and be really hard. So it's like everything is hard, which is awesome and fun. But like just getting it out there, getting the birds out there where they needed to be was like a huge pain in the ass. And yeah. And so and then we just spent a bunch of time like with those guys just yeah, just just showing them everything.

56:01And like also in the US, like we so we've been able to get around this by going on to like army bases. But other than that, like you can't fly like there's a bunch of rules about like heights and distances that you can fly at. And so we are mindful of that. And so it's like, OK, like we're going to we're going to send this thing 30 kilometers away at like, you know, 700 meters in the sky. And like, oh, OK, that works great. But like as you start to like push out, like, you know, there were there were a couple of times where out there was like, oh, shit, like we broke something like these guys, they still like have faith in us, but like, we need to turn this.

56:39And thankfully we were able to, but like, we need to turn this around like tonight, you know, like something wasn't working today. Like we need to show up tomorrow with an answer. And so that, those are like, I pulled a lot of all nighters like at this company, but like, those were kind of like the last batch of all nighters where it's like, Oh, okay. Like, yeah, this thing broke. Like we're staying up all night. We're fixing or showing up tomorrow and like explaining it to them. Um, and so that was, that was July. What we've been doing more of now is like, as much as I love to be in the field.

57:08And like, I guess what I'm trying to do is I'm trying to like hand that because we as a company, we need to do that. But like, I actually know pretty well how to do that. And also like how to teach that. And so that's something that like Sasha and I are both looking at like, handing off to people more and more. It's like, this is basically how you do forward deployed engineering. And like, this is how you go and like spend time with people and like you know just having faith I think too like we have a really good team that really gives a shit and they've been here long enough and understand what's going on well enough that like I can trust them to go and like to the spot and do the thing um and so I think like I'm spending a little bit like I'm spending a lot less of my time now on that and more on um what I what I've learned is this is that um you you do need to go there and you need to really like push it but then also every um because again we we sell like the product that we sell right now is like a component of a larger system and it requires engineering on behalf of the customer or the user and so you you inevitably reach a point where like you're handing it off to them and then you need to be there and be really responsive when they come back but it's like it's it's async um and so that's kind of what i'm looking at now is like how do we run as many threads as possible with as many different people as possible and then respond to those very quickly and then in those moments where it really matters to show up and like fix the thing or or it's a lot less of fixing the thing and a lot more of just like showing them like how to use it like go and do that but yeah I think sadly I will not be spending as much time like in the field anymore i think a lot about like reducing cognitive load and figuring out what is the thing in the company that i have to be doing for this it's the interview like i have to be the guy sitting in this chair uh doing this interview and everything else outside of that i should just basically try to figure out the best guy that can be an absolute tent at that one specific thing whether it's editing or like setting up out all the travel and all this stuff.

59:15For you, like, I know you started out as CTO, and now you're CEO. What has kind of that shift been like for figuring out what are the key things that you have to be doing versus finding people to do the other stuff? That's a really good question. This is my philosophy. I am not, you know, I haven't built a huge successful company yet. And I haven't, you know, so so this is just how I think about it. And I think that that's like, there's other ways that you can think about it that would be good too. Is just that like, if I haven't done something, then it's hard for me to know, like how to even talk to somebody else about it.

59:57And so I think like you can totally, you can and should hire experts that are way, way, way better than you at the thing. But to me, it's important to like go and do it first and get like an understanding of what's going on. So like whatever, and in doing that, then talk to a bunch of experts, bring in people like we've got really great people that help us with federal. Like we've got like two of the best people in the world that, that help us with that. But a lot of that like legwork is still me. And so that's kind of like an inversion where it's like, you've got, you know, you got like the steering wheel and then you got the engine.

1:00:36And so a good way to learn how to steer is just to like be in the car and to be that the engine and to do that. And so that's kind of how I look at it. It's like when we're first starting with the technical stuff, like asking a bunch of questions to a bunch of people trying to piece this together and then just go do it. Like you've got to be the one that writes the code. And then it's like, okay, like code's been written. That works pretty well. Let's find other people that, you know, okay. Testing, like I got it. Okay. Okay. I've done testing, like bring in somebody else. Okay. Got to go do integration.

1:01:04Got to do like this kind of more hand-to-hand sales. Okay. Know how to do that. Bring people in. And so now it's like, okay, how do we, how do we go up market? How do we do larger volume sales? How do we do sales to the DOD? Okay. I got a lot of like really great investors, advisors, and like people working for us that know those things way better than me. And like, I can ask a bunch of questions, but like, I'm going to be the one that's going on those calls and like asking those questions. And then And once I've got that mapped out, then you bring somebody in to actually do it. And then you just kind of like point and say like, hey, here are the things.

1:01:37And I think also like I'm still learning and trying to improve at like at the steering aspect and like what is the right amount to like, you know, how do you, okay, I know how to be an engine here, but like how do I steer other people or even like less steering? Like how do I just show them like where we're trying to go? and then they can mostly steer and and you are just you know it's a self-driving car um i don't know where this lane comes from but i think it's i've seen uh like ulu chang missouri yeah i think daniel strachman also say it where they say like if you want to basically you know build a thousand ships or whatever you shouldn't tell people what to do or how to build a ship you should just like make them yearn for the bath and then let's see um which i really i really like and it's a little like that where you're not necessarily telling everyone exactly what they should be doing and we're just like here's where we want to go here's why it matters yep yeah i think i think so too and and i think to me and and honestly maybe like i'll look back at this and and realize like you can't take more leaps of faith but to me it's like i want to if i want to build a thousand ships Like I want to build one ship myself just so that I know that a ship can be built.

1:02:50Like and and that I understand how to do it. And like I want to do that as quickly as possible. And then it's like, OK, the sea is great. Like, let's let's go to sea. Yeah, no, I think that's a great quote. So when you when you come back, you know, you've learned the initial stuff that you need to know to know how to do the thing. And then you can or you find someone that can do it really well. How often are you kind of coming back and basically not necessarily checking in, but making sure that your model of how to do it and is it being done well? How often are you updating that? Yeah. So, again, I think this is a place to like always be improving.

1:03:28one of the things like Sasha my co-founder actually said which I think was great is that like basically like evaluation is like one-tenth of the work of execution if that even and so I think yeah I'll just I'll just say this is like something I've really had to learn like being pretty young and this is like the farthest this isn't the first startup that I've started but it is the farthest that I've gone with a startup. And so really just to me, the biggest thing is like, oh, it is okay to provide feedback and to evaluate. And it's not even like, it's not even a harsh thing, but it's like, you just need to tell people like how they're doing and like where to improve.

1:04:11And so I think that that took me a little bit to kind of internalize is like, oh, okay. Like I actually, you know, I'm doing a disservice if I'm not providing feedback. And I just need to like come and look at this thing, whatever, like once every couple of days, once a week. And it's like, this is, you know, this, this, that, and it's, it's less of like micromanaging and, and really just like not being afraid to say what you think, even if it is potentially like this, not this sucks. I don't really see things like nobody's doing anything that sucks, but like, Hey, like, you know, just, just having kind of those, those difficult conversations.

1:04:50And I just think that that is, you know, I talked to any of my friends that are like, at similar or a little bit farther stages and startups. And it's like, that is just something that you really, like, kind of is hard and weird to learn. But I guess the thing for me is like, realizing that, like, when I have work for other people or something, it is like, yeah, like, you want your boss to tell you what to do, or you want them to tell you if you did a good job or a bad job like you do need to provide some some feedback loop yeah yeah so i think i think doing that and then also just the the place we've been really lucky is just that like i really feel like with the team we have like everybody just gives a shit and they just want to like go above and beyond and like make the thing work um and so they've been really good like i think as i've been learning that our team has been great at like helping even get that where they're like asking you know asking for feedback and just like doing you know they show things and and it's just yeah I think it's um kind of got got a bit off track there uh but yeah yeah you you're you're talking about like building a team or having a team of people that kind of give a shit how are you kind of trying to find those people and foster that I mean the fucking awesome thing about what we're doing is that it actually does matter and it is actually like a good thing for the world and like there are like well this part is not awesome it's horrible but like there are like good dudes getting killed every day and like we can actually prevent that um and so i think like from the top from the highest highest highest level it's like we're not starting with um i don't want to dog on any one vertical so the classic like fake you know like um uber eats for pets or something like it's It's not like we have some just bullshit thing that doesn't matter and we have to like invent some mission.

1:06:46It's like, no, like this, this is like people are dying right now. And if we don't figure this shit out, like our country is going to be fucked in the next war. Like we, we have no EW resilience. We haven't like our systems, like we, we are barely building drones in America. We haven't figured that out. And then when we do build those drones, we're building drones that were like hot shit five years ago. Like we need to, and that doesn't apply, like there's companies that that doesn't apply to, but like, we just need to really be like ahead of the curve on this. And so I think like from that highest level, like I give a shit about what we're doing.

1:07:26And so from that, it's just like trying to, trying to talk about that. And I think just being like even more, more and more, yeah like you don't really need to dress it up at all like it's just um it just it just matters and it's important and I think to me it's like very obviously like an inspiring mission I think to a lot of people it it is that too um and so I think the biggest thing is just not being afraid to talk about it and to be super vulnerable and talk about the actual real emotions of like, this fucking sucks. It's really scary. If you go to some of these places, it is scary. Like there will be like missiles hitting a couple blocks down from you.

1:08:18Like it is not fun. And that's okay. Because there are people that are in way worse spots than us. And like, it's not our job to get in a trench. It's not our job to get in a foxhole. But like, we should go as close as possible to that. And then help those people as much as possible. And yeah, man, like, I've never had to drive a car to like position and like there's a drone flying behind you at 140 kilometers an hour with an explosive on it and you're just hoping that like it crashes or they get jammed or something or you're just like gonna die um and so then it's like okay we're yeah like you you should go to the place where there is going to be a rocket that hits like the we're staying in a place and it's like you literally like are two blocks away from like just roof of the building gone and it's like that's really really really really fucking scary and also that's fine um and so I think that that yeah and and just to end this like I think I just see I think war is just the most awful thing ever and I there's this meme that I always comment on it when I see it on Twitter when it's like oh like men used to go to war and now they do xyz it's like well actually like you know having a matcha is way fucking better than like going to war.

1:09:49Like you don't like that's not there's nothing admirable about that. But what is also true is that like, people are at war right now. Like there there are hundreds of 1000s of people in Europe at war, and you can do something about it. And and like, it's um, yeah, it's just like, you you put yourself at more risk, you get into scarier situations. It's hard. It sucks. But like, also, there's there's literally hundreds of 1000s of people that have it not just way worse than you, but like, you can actually help them. It's not some like, oh, okay, this is some, some abstract thing that like you should worry about, you know, you can't you can't give a shit about everything.

1:10:30But like, if you do care about these problems, like, you can actually have an impact. And it starts with just a couple people. And then And from there, then it's more and more and more people. And like the beauty of technology and software and also just building cheap shit and trying to get your supply chain tight is that you can scale it. And so if you actually figure this out for those five dudes, then next thing you know, like, oh, now we're figuring out for 50 dudes. That feels a lot better than it did for five dudes. And we actually haven't had to do that much more work. and soon it will be for for 500 and etc etc etc um and yes i think just talking about it more and and just trying to put a lot of effort into capturing people's attention and finding the things that people care to hear about and and doubling down on those things um because because to me like anytime like somebody is thinking about or talking about the things that that we're working on, like, or, or adjacent things like that's, that's a good use of their time and thought.

1:11:44And that's a lot better than like, oh, what did, I'm dating myself here, but you know, what did Kim Kardashian eat for breakfast, right? Like, it just, it's, yeah. And so I think just like, we, you know, we have like a obligation. And I think you do a great job of this in like platforming a lot of different people, but like, you have an obligation to go out and like find things that matter and then talk about them. Um, because if you don't, I just spent so much of my life, like without a clear idea of like what I could actually work on that mattered. And just like, it wasn't in front of me of like, here's something I can have an impact on.

1:12:20And so then it's, it's this very aimless thing. And so I think just like realizing that the more you, you give, you give shape and, and like realize these things. Um, I just think that's really, really good i was i was listening to the christian you know podcast on the way over here and uh i wrote down this question that was basically uh biggest moments where you ran up against reality because you're building like hardware versus versus just software um what have been those biggest moments we are just like or you just really something up in a really real way um Um, yeah, I mean, we have had a couple of those moments.

1:13:06Um, I mean, I, I think we've had more of those moments than, than most people probably just cause, cause again, like that, that, oh, this needs to work 80 % or say it needs to work a hundred percent of the time, not 80%, 90%. Like that wasn't something I came into this knowing, right? like I I'm I I don't have uh I just figure things out as I go along and so um that comes from like fucking it up 10 % of its time but those that 10 % was like really big and shitty and so um yeah I don't know how like good of a light this this like we've just yeah we've just really fucked up like the first time we went to Ukraine we spent like a week um doing this integration Like I was, I've gotten over it since, but like just really scared to be there and just worried and worried by the air raids and, um, trying to, we wrote our whole code base in Python to begin with, which is really stupid.

1:14:01And I had to rewrite the whole thing in C++ cause like, it's like, I'm not going to ship my source code out to people. But so I'm, I'm doing that, like not sleeping, just like really stressed. And we spent all this time like integrating with these guys and they had, um, we got a catapult over there. The reason we have a catapult over there is because we went out. These guys had a catapult and like everybody in Ukraine has a catapult. They all like their drones off catapults because you can you can like kind of like the zipline thing where it just shoots. Yeah, it's just like, yeah. And, you know, the reason just being that like landing gear is bad for aerodynamics.

1:14:36Like basically the way to get the tightest aerodynamics on the plane is to design it such that it's already in the air and you don't have to worry about getting it up. like like with VTOL or landing gear or something and then if you design it so that it's already in the air then you got to get it up in the air so you get this big pneumatic catapult that just shoots the thing out and then and then it takes off but so yeah we just weren't tracking that and like when when you're building like just with with this like the the takeoff is like a pretty important or used to be, should I say, a very important part of, of flight.

1:15:13And, um, there's, there's inertial measurement units in here. And basically there's, these are, these are complicated systems. And so it's important that you test with, um, basically understand all the variables. And, and so, especially when we were starting like takeoff was an important variable. And so now we just get out there to this field, um, we see this catapult and it's like, Oh fuck, man. Like this is this is our first flight that we ever did um there and we're looking at the catapult and this is something i've gotten a lot better at is like saying no when shit is not what you expected but only from saying yes to too many things and and then it doesn't go well but so yeah we put it on the catapult um and normally we have like uh here when we do testing definitely and and with most people um like actually less and less but um we have like telemetry feeds so you're getting information off the drone about what's going on.

1:16:04You can just kind of see the status of everything you can send, like, you can see if something's going wrong and say, like, come back or do that, you know, you have a lot more control. And also, we don't do this anymore. Well, we do it in the US, but like, you know, having like a backup GPS. And I can go into that in a second, like, because that's, it's, it's a whole own thing. But like, yeah, we get out to this field, um haven't been sleeping for like a week just like really scared there's this catapult and we just like yeet the drone off the catapult and it's like oh like yeah we just don't support like a catapult like it doesn't like and so the drone like kind of flies around it like looks pretty good but like we can't tell what's going on and so it like looks pretty good and it's like flying flying flying doing a circle um and like it wasn't working it's just like you know you look at the logs you fit this stuff out but it looks it looks pretty good and then all of a sudden it just like takes off like completely away from us just takes off away from us and it's like going going going and like we don't have any control over the thing only one guy does and like going going going on and then finally he like flicks the thing we're just kind of sitting there and we're like well shit like guess the drone's gone um first flight in ukraine and and and then like two minutes later and thank you Artie Pilot.

1:17:22It just like came back, you know, it went on to like on these short distances, like you can totally use, like our system is really useful for like hours long, 100 kilometer long flights. Like if you're a couple hundred meters away, like you can just get back with a compass and like an accelerometer. And so it came back and landed, but that was like a big, like, oh shit, like that sucked. and now we got to go look at this and figure all these things out. The way we've gotten away from having those moments is just, yeah, just testing all the things ahead of time. And then if something doesn't, if something is unfamiliar, like not to say we can test everything, but like, we'll figure it out really quickly.

1:18:09Like we've got really good engineers. We can test stuff. Like we've definitely learned a lot. Like we're out in, we were out in Dnipro and like our, our system is waterproof. But we had done some like crazy, we're always, we have always like, we're very willing to do like crazy modifications. So, so for whatever reason, like we've, we've popped it out of its like waterproof case just to mess with some different stuff. And then we've like put it on top of the drone. So we've got this like exposed board on the top. sensors underneath there was a good reason for this um but then it starts it starts raining a little bit like drizzling yeah just just drizzling and you've just got this like exposed and i i'm like kind of an idiot man and and like all the guys are like just fucking send it i'm like yeah dude like let's just go let's go let's go and sasha just looks at me and he's like man like no like it's just not a good idea and that's where it's like oh yes like let me let me back out for a second and see like and then we just waited for 10 minutes and then we just did the test with again like you can fly this in the rain but you can't fly this like hacked and so i think that that was like a really um but that that just comes from like again like to the thing i said about like everything is like your gut um just training the gut to oh okay i've been in this spot before i've wanted to impress these guys who are fucking badass like these are the coolest people in the world i really want to show them this thing but let me take like a step back and and actually impress them by not uh or making sure that it actually works or giving you the best likelihood of working after you've gone through kind of like the checklist of things to lower the likelihood of failure yep yep and just having having a checklist yeah did you uh did you like previously not have a checklist at all and you were kind of like again going on instinct but before you train the instinct you don't really you're you're going on gut but you don't have the feedback or like you don't have the prior information this is the funny thing is like every process has to be like invented so so we have like a really good checklist we've got like pre-packed like we've got literally pelicans that are packed with all the shit that you need to like go.

1:20:27So like, you don't have to come in on Sunday for your flight and like pack a bag and then you forget something. But like all of that just comes from, from, yeah, like you're doing, when we started, we're doing like 90 things at once and we're trying to figure all these different stuff out. And so a lot of that process like just came. And that's been another thing that like growing the team has really helped with because you just have to have process. And also a lot of the people, um a lot of the people that we've hired are just like really good at that and just very thorough and so yeah to begin with just a lot of vibes and just forgetting stuff and then being like oh this is really stupid i know that for surgeons they have a very like clear process before going into a procedure um and before that there was people just like not washing their hands and stuff it You know, you go from like a surgery to like delivering a baby.

1:21:21My dad's a pilot. And like growing up, I would, you know, always always riding his plane. He's very thorough. Like any pilot knows this. Like you have you have the checklist. You go through the checklist. You check everything like that. That's what gets you killed is not checking your checklist. And that's just one of the big things that we have had to internalize and do is just, yeah you're you're doing serious shit with serious people and you need to behave seriously and um and it's really great and awesome to be like uh young and figure things out quickly and be kind of a goober but also like you you got to get your shit from the past and make sure that you're not just like unnecessarily taking bad actions just because you just don't know just these people like when you deal with really competent people and i would say that like anybody in especially us socom but like globally like a special forces soft operator is like these guys are fucking elite and and they're not used to people fucking around and so they will give you a lot of like trust and credit but like you can't you can't abuse that and like you you just gotta you gotta show up knowing your shit because if you say um if you say that this thing does X, Y, Z, like they just trust you.

1:22:36They don't have time to verify that. You're basically every interaction you're investing a lot of reputational capital. And if, you know, if it works great and, you know, it comes back, if it doesn't, you've lost a lot. Yeah. Yeah. And I mean, we, there's really just one, like there's only one such interaction that haunts me, but it's like, it's like, yeah, you just, you get the opportunity to work with the best people in the world. and you're just not on, like literally the best people in the world. And if you're just not on your shit and doing what needs to be done, then you put everybody in a bad situation and then they're just like, hey man, this ain't it.

1:23:16And so I think that's been just a big, like I love thinking and highlighting. And like, honestly, a lot of this has been talking about things that like weren't wins. And there are a lot of wins, but I think you just you you learn a lot more from the losses and like that's definitely like much more painful yeah oh yeah yeah much much more painful and just like yeah you just uh I think you just grow up a lot when when you have those moments and it's like oh shit man like I I'm sorry how has like did you grow up being willing to take risk and willing to kind of push through and put yourselves in situations like that?

1:23:58And how has that kind of changed from the first trip to like deciding to do that in the first place and deciding to go back? Yeah, I think growing up, I was, yeah, I just, I wouldn't even say it. Like, I think in the past couple of years and kind of the zeitgeist of San Francisco has shown that like, yeah, there's these returns to risk and you should seek out more risky things, yada, yada, yada. There's this good model. I don't think that that was like exactly the model that I had growing up. I think it was just more like being really interested in certain things and then really pursuing those things.

1:24:39And so some of those things were like super like normal and mainstream, like, you know, okay, like really, really into basketball. Like I really care about this. Like, you know there's not um there's there's really nothing like outlier about about being super into basketball um but then sometimes I just get really into things that that I think were like weird or like out of distribution at the time um yeah many examples of that yeah so I think the first like really crazy thing I did and and for a long time this was a thing that I was like most proud of um was when I was like entering my senior year of of high school so like when I was 17 I went on the Appalachian Trail by, so I started off with one friend and then I ended up by myself for, for like a month.

1:25:26And I hiked like 400 miles of the Appalachian Trail by myself at 17. And I literally just like read a Reddit post where somebody had mentioned that. And I was like, that sounds really cool. Like, I want to, I want to do that. And that was like all the thought, like there was no more thought. It was just like, that sounds really cool. I'm going to, go and do that slowly got a cult adventure and you're like I'm gonna go do that yeah yeah and and I think that like I I just had really supportive um parents and like supportive in the sense of of like giving me the things that I needed to to do that and then also supportive in that like I think they realized at a pretty young age that like I was kind of just gonna do the things that I wanted to do and so they should just like facilitate that um and so So yeah, that was like definitely pretty wild.

1:26:18And I think shaped a lot of my adolescence afterwards because it's like, yeah, like you're literally sleeping, like you're either sleeping out in the woods by yourself or you're like sleeping like under benches and you're like hitchhiking and doing all this crazy stuff. And what I also really liked about it is that it's like a very clear goal. So it's like, you know, anybody that's hiked some or all of the AT is like you just follow like the white flags like it's a trail it's literally you just put one foot in front of the other um and so I think that was very nice just like goal and every day you you do that and and do a little more of that um and yeah just so I think it's less about maybe about risk and and just more that like um there's there's really interesting things and some of those some of the most interesting things require risk how many trips to Ukraine have you them um i've only been there three times three times yeah okay the the first trip was like a little over a week the second trip was like a month and then this last trip was a month yeah and so yeah i uh what we're doing now is like sasha and i are like one of us will be there and then we'll tag the other person out and then we'll we'll kind of flip-flop there uh because to me it's like we just we need to have what what became clear is like everything needs to work really well and be really tight before you go over there because there's just no appetite for stuff that doesn't work and it's like you don't want to waste anybody's time and you you just want to show up with shit that works but uh once you have that like you need to have somebody there like in country like we're gonna be there until like until the war is over and then after that so that's gonna be like a long time how was that first trip how was the second trip and kind of what's the like feeling that you have when you're going over there yeah yeah there's a lot less action the first time we were there to be honest i just don't think i was like in the right head space so like i said i was i was basically coding like the whole time and just like not sleeping very much uh which is not good for uh i think critical thinking is like being uh in you know fight or flight comfortable yeah and um and i think also just like i just really feel like um it is it is a really it's still intense but it's a good feeling to go and you have something that you think is like really going to be helpful to somebody and the first time we went over there it was just like yeah we're meeting people but i just i don't know man like there's a lot of companies that go over with shit that doesn't work and that just want to like they just want to use this whole thing as like this marketing thing.

1:29:04And it's just, it just, it doesn't, to me, it didn't feel good to be there, like talking to these people and having like something that I wasn't confident in because yeah, we're able to work with them. And I think it was a great learning experience, but it's just not, I just don't think it's fair. And I think that there's like this really awesome like silicon valley mindset that that just doesn't apply in this case of like oh you should you should like sell the thing and then do it like it's exactly flipped it's you should it's just not this isn't the kind of shit that you want to sell before you have it because like timelines are important and like there is a lot of risk and like yeah if you sell somebody like this um you know enterprise sass thing and it doesn't come through like their enterprise is still running like but but it's just it's just a very different thing and and so I think the second time like yeah we we we got into significantly more sketchy situations but a we were we were with really good people that are also just like like I said like we you know I do my best but like I'm not some like I'm not a soldier right and I've never been a soldier and And so when you are with those people and they have like an understanding of risk and a very different model.

1:30:26And it's just like. They've been in these situations and they've actually stayed around. Yeah. And they just talk to you earnestly. And it's like, dude, I'm in the risk. Like we talk. It's crazy to us as like people from like the best country in the world where things, you know, nothing happens to you when you're sitting in your house. But like, I don't know when when you go out and you're with guys that like do actually go into a basement and fly drones for a week at a time. And that is like super sketchy. And they're just like, yeah, like there's missiles that hit the city. You're going to be fine.

1:31:01I mean, nobody's it's just like if, you know, if it's your day, it's your day. And that's just like it's just really the mentality over there. Yeah. Oh, yeah. I mean, well, it just it just has to be right. Like it's because you just got it. I don't know. I think I've learned so much that like it just comes very naturally to people that have actually been in these situations. But I've just had to hear it. But it's like, yeah, you just you you get out there and you're just you just can't. I don't think it really does well to be worried about those things, because, again, there's like nothing. and and and for me too and like even this is such a it's like we we purposefully don't get ourselves into the sketchiest possible situations and like it's just not that's not like how i see our job and so it's just like there's people that are taking a hundred plus times more risk than you thousand ten thousand times more risk and so i mean once you start actually interacting with those people it's it's just um it's like yeah i can do this it'll be fine do the people that kind of come over there with a half-baked product have the same sort of mentality that you have being willing to do that if it's if it's the right thing to be doing i think the independent variable there is like looking for the single success versus looking for like continued success and looking to eliminate failure because from what i've i've just learned about like yeah there's you you can go over there and you can get the picture you know you get the picture with the guys you do the one thing you put it on your website you talk about it endlessly and it's like one it's like oh we we have we've done this thing like i there's a company that comes to mind it's like yeah we have this you know we did this strike you know here's this whole we're going to talk about this endlessly and then you just leave because you don't like again failure sucks and like if you just if your whole thing is like we want to do this once and talk about it um and you fail a hundred times you do it once and you leave like that's shitty whereas like to me it's like i want to be there until it literally just works like the only reason i'm not there right now is because like it actually works without me being there and we still have people there and it just needs to be like and and this is the case with any customer but like it's the case in ukraine as the case in the US, it's like it just needs to it needs to not break.

1:33:28It needs to never break. And so your mentality is you're going over there to basically solve the problem and make something work, whereas some other people are going over there for like a photo shoot. I think I think that you see the other thing, like the thing I'd say is most of the people that we interact with at this point that are still there, like, dude, it's been three and a half years. Like we were really late to this problem. Like we were not this specific problem, but like we're late to the war. Like it took us a long time to get our shit right and even get over there. And so I just think most of the people that have stuck around and are still out there like do give a shit.

1:34:05And I think like 90, like I think almost everybody in the world is like really good. And I think most everybody you run into out there like really does give a shit um and and i think most of the um the the war tourists have kind of gotten disenfranchised and like i can also say like having having done it's like i i know what it's like to go over there you do the one thing that you can talk about and then a bunch of stuff doesn't work the thing is you have to like the the i think to me the difference is like do you figure it out and then you come back and you say like let you say like here you go like and you say I'm sorry and you continue to like make the thing work like and and and do that and so um and there's not like that's that to me there's there's no like it's not a financial thing it's just like it needs to I think a lot of the financial incentives push you away from doing that but it's just like you need to go and do that um at least at least that that's that's kind of the differentiator to me is just like, I have gone there, I have failed, I have failed, and then left.

1:35:20And then I've come back and made things work. And and and that is on. I think that second part is like, all you can control because you always will, you always will fail. But like, it's kind of what you do after that, that matters. I think people can kind of sense when you genuinely care, and that you're genuinely there to like, make something work or make something happen. um yeah also thing they can genuinely they know when that's not the case dude and and like i will say like i you know it to me it's like a honestly an honor but like you look somebody in the eyes and it's like this this commander and you're just like let me tell you like i have been working on this every single day for 15 months i have this is the only thing i care about i have built this whole thing for you these are all the things that it does it is like yeah this question that you have like i figured it out because somebody like this is specifically designed to help you and i think that like that is um that's just a really it's also honestly kind of like a rare thing that you can kind of bring to someone and and show them yeah yeah it's um i mean i think i think in the drone space, there's, there's people that, um, started before the war in Ukraine and after the war in Ukraine.

1:36:44And, and, and there's like differences and there's definitely like a lot of really good technical expertise. Um, and I would say even the folks that were doing it before the war have more technical expertise. Um, but I think there's just kind of this like drive of like, I, you know, if this wasn't the thing and I could do some other thing, I would do the other thing. Um, it's just, you know, this, this is what it is. And, and so, um, I, I do think, I do think that that matters, uh, to folks. And I think also just like, none of this shit is easy. And like, when you still show up and you do, you explain like all of the, I don't know, it's, it's not like a, it's just like, yeah, this was hard.

1:37:30That's fine. Like we've done it. Like, let's, you know, let's, let's go. You've clearly got a very, I think, I think like a pretty solid intuition on just making things happen. Even if it's like kind of a duct tape solution, going from that to something where you are trying to hit nine nines of reliability is kind of like it's not the exact opposite, but it is kind of the exact opposite. And what is the flow in your mind and the way that you iterate from that first iteration to getting this thing? Dude, it's such a good question. So So a couple of things. I think oftentimes the hardest thing about anything is plugging everything in and turning it on.

1:38:10And so like I think about everything is like kind of like either black boxes that are connected with like, you know, data pipes or just like functions basically. And so you the most important thing is like mapping out how the black boxes all connect to each other and then you have to fill them in. and you need to do better and better and better. But so I think that was like the really nice thing with like going and just doing the flight is like, okay, this clearly doesn't work. We're not expecting it to work, but we do know how we pass data in. We do know how all these systems like connect and we have an understanding.

1:38:47So like, I think probably, yeah. I mean, literally every single part of that has been like swapped out and modified, but you can find bit by bit by bit, like what is the thing that's most broken here And how do I swap that out? And so then I think always finding the thing that's, that's most broken and swapping that out. Eventually you reach a point where actually things aren't nearly that broken. So like, if you break it down into like three stages of like, okay, you have that, that thing that's like clearly ridiculous. Then you, you fast forward like six months and you have this thing that like kind of mostly works like 80, 90 % of the time.

1:39:22Maybe you have to like SSH into it and like type a little bit on the laptop, but like it will generally go up in the air, but then sometimes something goes like catastrophically wrong. Like clearly that is like significantly less broken than the thing that like doesn't even fly at all. And you just get there by like unfucking all these different parts of it. And then I think from there, then like you just kind of repeat this process. I've repeated this process iteratively. And then you reach a point where like, A, it becomes clear. It also comes down to like requirements too. Like initially, like we weren't our own customer, but we were the person that we're trying to deliver on.

1:40:03So we needed like a prototype and we needed these requirements were like much less stringent. We just really needed to show that like we could get a drone up and send data into it. And like that there was some, you know, we're collecting something and we are able to localize off of it, yada, yada, yada. Then the requirements become more stringent because if you go to somebody and you show them something and you say, hey, I think this is pretty good. Or I think this isn't great, but I think this is like directionally correct. And then they say, that's great. Like do X, Y, Z. Now you have this thing to build against.

1:40:35And then you can start to see what's really broken and like keeping you from that. And then you reach a point and like, we just totally reach this of like, oh yeah, like we want to sell a hundred of these a month, like this month, you know, like of course we want to sell millions of these, but like this month, like we're selling 100 of these that's that has a different kind of like requirement because because a that means like you need customers that are buying more than like the onesie twosies so you need repeat customers so you need people that they make something so good that they want to come back they get two of them and they're like this is fucking awesome give me 20 and then from 20 this is fucking awesome or the product is so well made and like you know i think to get from 20 to 200 to 2000 then you you start having like, it's not just questions of like quality, but it's like, how are you building your product roadmap?

1:41:26And like, do you, how well do you understand your customer and their needs? And like, what are you providing to them? Because like, obviously just replacing the GPS is like, cool, but like, there's a lot more on a drone than that. There's a lot more you can do with software than that. And so just, just building more and more of those things out. But I think it just comes from, you know, like, I think your initial question was really good. It's just like, I just want to make something that works well for the people that need it and it eventually became really clear that like uh they're happy with the hacky stuff but it just needs to work well and so and so you just go bit by bit by bit yeah show that you can move fast but then also like on the back end make sure that they are so happy that they're coming back again yeah and everybody I mean we've um we've sponsored like different hackathons and like university projects and stuff and basically like asked and encouraged people to like build the exact thing that we've built and like I I think that that's great and and you can just totally see when somebody has like inklings of something awesome or like is doing like there's just such a big gap between production ready shit and like an idea but people still appreciate if you take like nobody's expecting you to actually build the whole thing that's like, you know, TRL nine in a week.

1:42:48But if you show them something in a week, they're like, oh, okay, you're taking this seriously. Like, let me invest into you. And I think that's also another big part is like, I think we've just luckily continuously demonstrated to people that are like really, really good and competent. Like we give a shit, we're taking this seriously. Like a, if you see something that's, that's can be improved with what we're doing, like, please tell me and I will improve it and work really hard. And then like, yeah, we give a shit and are doing everything we can to make this good. And so I think that just leads to investors, but more so like customers and like people that are just willing to take a bet on that because, yeah, because it, Rome wasn't built in a day.

1:43:35Yeah. People understand that it just takes time. Yeah. Yeah. and actually like people starting the fee zero on going to alt weight to like v10 yeah yeah no 100 and i mean yeah when you start like when you start at a hackathon it's just like well clearly you're not building like you know the yeah you're not building a tesla in in 24 hours you might be building the like clay version of something that you know uh the silhouette of a core. Yeah, exactly. If you made like a whole clay Tesla in 24 hours, people would be really fucking impressed. Final question. What are you most excited about? Yeah.

1:44:13What am I most excited about? I think that it is really, really fun and awesome to go from doing like a handful of things. Like, I think it's funny when people self describe as like, oh, we went from zero to one. I don't think you ever go from zero to, you do one thing and then like, okay, maybe if you - Make one little tweak, you make another. Or also like, okay, meta, they've probably gone zero to one as a company and then they paid off all their investors. They're still doing new things. I think you're constantly resetting that. But like, I think going from let's just make this thing work at all to let's make this thing work really reliably to let's make this thing work reliably at scale.

1:45:02And then let's make this thing work reliably at scale on board tons and tons of platforms. And so there's so many different drones out there. like the word drone is such a catch-all and it can describe something that's like you know literally like a tiny whoop like 65 millimeters all the way up to like massive like thousand pound like group five like like combatant like uh they cost like five dollars up to something they cost like 20 or 50 million yeah and and um and i just want to go on as many of those as possible. And I think in order to do that, you need to reconfigure parts of what you're working on and you do need to do engineering.

1:45:45But like that's what's really exciting. It's like, okay, we've got this nice box, but really this is like, it's kind of like a developer kit. Like you can put this onto some amount of drones and you can use it and you can see how great it is. But like, Like I'm excited to work with the people that are delivering on scale on drones to deliver them capabilities on those drones at scale. And so that to me is very exciting because I think like, yeah, software scales, like there's a beauty to scaling software, but like it does scale pretty naturally. Like I, you know, random projects, like work thing that I did.

1:46:25It's like, yeah, we made this thing spin up, spin up a database in literally in like 30 minutes, send it out. And oh, my God, like it's got, you know, 1.5 million hours of watch time. There's all these rows. It's like it just scales naturally. Like it's super easy. And like literally like I hadn't touched the thing. It's like, holy shit, all these people are using this. And then you go from that to this. And it's like, yeah, it's very there's it's very much like hand to hand combat to scale like this. and you need to update the product, you need to do engineering, you need to manage supply chains, you need to manage like a lot of aspects of what you're doing.

1:47:04And that's fun because it's like this fun new challenge and it's kind of like a superset of the thing that we've solved already. And so, yeah, that gets me like super jazzed is just, yeah, I wanna go from hundreds to millions. and and i just think that that is like that is the mountain is the exciting yeah yeah and it's very binary like it's um yeah okay you you have you have the box that that works at at uh smaller scales like now get the box that works at scale millions like that is uh that that's a binary and it's it's like a bit that we haven't flipped yet and so i'm i'm very excited about that

From the publisher

Flying drones in GPS-denied environments.

#49 - Ian Laffey, Cofounder & CEO of Theseus

Ask about this episode

In short

Transcript

More from Relentless