EPISODE 1880

[INTRODUCTION]

[0:00:00] ANNOUNCER: Simon Shuster is a journalist who has reported on Russia and Ukraine for over 15 years, most of that time as a staff correspondent for TIME magazine. He was born in Moscow, and he and his family came to the United States as refugees from the Soviet Union when he was 6 years old. After graduating from Stanford University in 2005, Simon returned to Moscow to work as a reporter for the Moscow Times, Reuters, the Associated Press, and other publications. His political coverage of Russia's descent into authoritarianism included numerous profiles of Vladimir Putin and interviews with Dmitri Medvedev and other top Russian officials. He has also interviewed and profiled the last three presidents of Ukraine, starting with Victor Yanukovych, whose violent overthrow in 2014 he covered from Independent Square in Kyiv. 

That winter, Simon was the first foreign reporter to arrive in Crimea as it was occupied by Russian troops. Since then, he has spent years covering the war in Ukraine from both sides of the front line. The year after the annexation of Crimea, Russian authorities deemed Simon a security threat and banned him from entering the country. 

Simon first interviewed and profiled Volodymyr Zelenskyy in the spring of 2019 when the actor and comedian was in the middle of his long-shot campaign for the presidency. Since Zelenskyy's election victory, Simon has been granted unparalleled access to his administration, his close friends, aids, and staffers. He has traveled three times with President Zelenskyy to the front war in Ukraine and has spent months reporting from inside the presidential compound in Kyiv as the Russian invasion unfolded. Simon is the author of the 2024 book, The Showman: Inside the Invasion that Shook the World and Made a Leader of Volodymyr Zelenskyy. And he recently wrote the TIME magazine article, 'How Ukraine Gamified Drone Warfare'. 

He is currently at work on a new book that examines the future of warfare and how the lessons and technologies that emerged from the war in Ukraine are changing warfare and security around the world. Simon joins the podcast with Kevin Ball to discuss drone warfare, AI-assisted targeting, the gamification of drone combat, the rapid iteration cycle of drone innovation, new ethical dilemmas in warfare, the coming proliferation of war drones, and the shifting balance of global power. 

Kevin Ball, or KBall, is the Vice President of Engineering at Mento and an independent coach for engineers and engineering leaders. He co-founded and served as CTO for two companies, founded the San Diego JavaScript Meetup, and organizes the AI in Action Discussion Group through Latent Space. Check out the show notes to follow KBall on Twitter or LinkedIn, or visit his website, kball.llc.

[INTERVIEW]

[0:02:58] KB: Hello, and welcome to this episode of Software Engineering Daily. I'm KBall. I'm your host today. And it is my extreme honor to bring a somewhat unusual for our topic area guest on. Let me introduce senior TIME correspondent, Simon Shuster. 

[0:03:14] EC: Thank you. Really great to join you for this conversation. Thank you for inviting me. 

[0:03:18] KB: Yeah, I am very excited. So, let's maybe start with a little bit of your background and what you've been deep diving in for the last, I don't know, how many years. 

[0:03:26] SS: Yeah, for most of my career, which is about two decades now, I've been reporting primarily on Russia and Ukraine. Covered the war in Ukraine since it began way back in 2014 with the annexation of the Crimean region. And basically throughout, I've focused on it interviewing military commanders, presidents of Ukraine, people on the Russian side. And the last 3 and 1/2 years or so since the full-scale invasion began in February 2022, I focused very intently on the war, especially President Volodymyr Zelenskyy. I wrote a biography of him called The Showman based on my extensive interviews with him in TIME embedded with his team, especially in 2022. And I'm working on another book now about the future of warfare and how the lessons and technologies that emerged from the war in Ukraine changing the way that wars are fought and security is thought about around the world. 

[0:04:27] KB: Yeah. Well, and I think that is a great intro to why we're having this conversation, which is this war has really been a very technologically intensive and shifting war. And there's a lot of really interesting stuff from a pure technology standpoint that's been coming out of it. Do you want to, maybe looking at that last few years for folks who haven't been paying as close of attention, give us the high-level of how warfare has evolved in this conflict? 

[0:04:54] SS: Well, this war began as very much an old school battle of tanks and artillery and infantry. Armored vehicles facing off and shooting each other from various calibers of cannon. Certainly, aviation was involved, especially on the Russian side. Russia had a great deal of air superiority for at least the opening days of the war. But very quickly, the Ukrainians, their society mobilized. And what you began to see was a great deal of innovation happening with assistance, I should say, from the United States and the European allies, but really a lot of it was homegrown. A lot of it was coming from literally people, civilians tinkering in their garages during the invasion and figuring out ways that they could use technology to fight back, to cancel out Russia's numerical superiority, its superiority in kinds of weapons that it had, its air dominance. 

And that's what I followed very closely. I did a lot of my reporting looking at who were the people doing this tinkering. What kinds of technologies helped them balance out the war, defend themselves, and, in many ways, start to turn the tables on Russia? What we've seen, again, very high-level, kind of a three-and-a-half-year story, and it's still ongoing. But what we've seen in the past year, certainly, I would say even maybe more than a year, is Ukraine finding high-tech ways to do deep strikes into Russia that really hurt the Russian economy and the Russian war machine. 

In the last few months, the Ukrainians have unleashed incredibly effective drone strikes against particularly Russian oil refineries and other energy infrastructure that have caused something of a gasoline shortage in Russia. So all of that adds up to technology helping Ukraine tip the scales in this war in a way that really no military analyst, strategist, politician expected at the outset. 

[0:06:56] KB: Well, and there's a lot of different pieces of this, but let's maybe follow the thread that you started there of looking at drones. How have the Ukrainians, in particular, but I think also the Russians, shifted their warfare to utilize drones? What's the actual hardware being used? What are the different pieces that go into that? 

[0:07:17] SS: Yeah, it's a big topic. We could definitely spend a whole hour on this one question. But early on, the Ukrainians were using very small-scale FPV drones. So, first-person view drones that are generally designed for drone racing. To some extent, maybe in agriculture to monitor crops, but also just for hobbyists, really, to take wedding photos and things like that. 

These drones were available. The military in Ukraine was using them for reconnaissance and some other tasks before the full-scale invasion. But with the full-scale invasion, the Ukrainians really began to see the value of drones like this on the battlefield, especially when it came to being able to see the enemy, see where the Russians were advancing, and respond with forces with other kinds of weapons. 

The kinds of drones that they were using primarily were ones that were available on the commercial market. The Chinese were and remain dominant in this industry. A company called DJI is a Chinese company that, years before the invasion, really established itself as the world's preeminent maker of FPV drones and the various components that go into them. So these things began flooding into Ukraine, often purchased by volunteers, donors, people who were buying them up on the market in Europe and the United States, and shipping them in mass to Ukrainians. 

Once in Ukraine, they were either used for reconnaissance using the cameras that are already affixed to the drone. Or very quickly, the Ukrainians began doctoring them or changing them to be able to carry weapons of various kinds, either as what's kind of popularly known as kamikaze drones. So, a drone that is fixed with an explosive, flies into its target and explodes. Kind of one-way drones or drones that are capable of dropping explosives. These technologies sort of iterated quite quickly and began to advance in many different directions, where then you had sort of even naval drones that were basically small boats that were remotely controlled. 

One other point I'd say, just kind of in this context, is the importance of Starlink, is I think you know something that we really have to note early on in this conversation. One of the key innovations, which now seems quite obvious in hindsight, but when it happened, according to my reporting, this was in the spring of 2022, really within a couple of months of the full-scale invasion, the Ukrainians figured out, "Hey, we can fly these things much farther and really secure the link between the pilot and the drone if we put a Starlink antenna right on the drone itself. And once they figured that out, I actually met the team that did that and visited the place where they first tried this thing out. It was really in a garage on the outskirts of Kyiv. They put this thing on. They attached the Starlink to a drone and quickly realized that this was really a groundbreaking innovation. Starlinks became and in many ways remain essential to the Ukrainian drone program. 

[0:10:28] KB: That's wild that it happened literally in someone's garage. The revenge of the tinkerer. 

[0:10:34] SS: Yes. Totally. Totally. I mean, there's war in so many ways. Sorry to interrupt, but it is like the revenge of the nerds on an incredible world historical scale. The people who have been at the forefront of this kind of warfare that has really worked to Ukraine's advantage and changed the nature of the conflict have, in many cases, been geeks, engineers, computer programmers who were mobilized into the war effort early on. And as military commanders, both NATO commanders and Ukrainian ones told me, when you're fighting a regular war, you often have the kind of regular army to work with. It's the people you've recruited, trained. 

Whereas, when you have a full-scale invasion like the one you had in Ukraine, the entire society is mobilized. The entire society begins looking - every individual begins looking for ways to use the skills they have to serve the national defense. And what you get is many people have many different skills. Some of them are very good at cooking soup, and they cook delicious soup for the soldiers. Others are very good at programming, and they begin to program and design weapons. That's what we saw. And I think that helps explain a lot of the speed and agility of the development of these weapons in Ukraine. 

[0:11:53] KB: Yeah, absolutely. I want to dig in more on that. But before we do, since you brought up Starlink, I feel like one of the things that I've kind of seen here is this arms race in terms of communication and jamming techniques as well. And how are we able to talk to the drones at a distance? How are we able to interfere with our opponent's abilities? Can you maybe talk a little bit about what that progression has been? 

[0:12:15] SS: Yeah, it's critical to understanding the challenges of drone warfare. So, if you're working with drones like the Chinese ones, the FPV drones, Mavics, they're called, that are designed for racing or wedding photography, those drones are not really made to operate in a military environment in a battlefield. In a battlefield, you have electronic jamming where one or both sides are using electronic warfare to suppress or block radio frequencies. All of these drones that are available on the market for anyone to buy, they use radio signals to communicate between the controller and the drone, the pilot and the drone. 

But in a war zone, you have all kinds of jamming techniques that either cut off that radio frequency. Or you have different techniques like spoofing and other things that allow your enemy to actually seize control electronically of the drone and switch it to an enemy's controller, so that then your enemy is in control of the drone that you just sent up to fight the war. That becomes the kind of, I'd say, at least one of the defining challenges, if not the defining challenge, how to get around, how to secure that link, that electronic link between the pilot and the drone. And that's when things get really inventive and strange. 

Starlink was probably the most effective, most famous way around that. If you have a Starlink connection, that means you're controlling your drone basically through the internet. It's not using a radio frequency. So Russians jamming the radio signals won't jam a drone equipped with Starlink because that drone is communicating through Starlink satellites, SpaceX satellites, through the internet to the controller. That's a great solution. 

Over time, and what we saw over the course of the conflict is both sides figuring out new ways to jam each other and use different creative ways of electronic warfare. It's called EW, a term you often hear in this context, EW. One of the funny solutions, I mean funny in that it's just so bizarre, is, for example, the Russians and Ukrainians have started in the last couple of years using drones that fly on fiber optic cables. You have literally a drone that has attached to its butt a spool of fiber optic cable, and it's unfurling as it flies. 

So, it's attached physically to the pilot through this fiber optic cable, and that gives you a very clear signal. And it can't be jammed. It can't be blocked. But, as you can imagine, that comes with all kinds of other downsides, like your spool of thread can be caught up in a tree, cut for some reason, or things like that. So both sides got very creative and began iterating and really racing each other in this incredibly fast cycle of innovation, especially in the field, not only, but especially, I would say, in the field of electronic warfare, how to jam the enemy's drones. 

[0:15:22] KB: Can we maybe talk then about - each side is doing this sort of tinkering, exploration. Revenge of the nerds is great. I think all of our audience can identify that's what we like to hear in a lot of ways. But what are the toolkits people are using to modify? Are they using off-the-shelf parts you could buy commercially? Are they doing modern stuff? What type of software is involved? How is this innovation loop happening on the ground? 

[0:15:47] SS: Primarily, they're using off-the-shelf stuff, circuit boards, chips that you can buy on the market. It does start to get more sophisticated over time, where you have, I'd say, more in the field of kind of command and control systems. So there's a system that the Ukrainians developed with help from American private companies and the American government and military, a system called Delta. That is not an off-the-shelf system. That is a very complicated system for basically battlefield awareness, where you have interactive map of the battlefield showing you, for example, where all the drones are flying at any given time. 

One demonstration that I saw of this system, you can literally click on a drone or a quadrant of the battlefield and click on the drone and then see what the drone sees on your screen of the Delta system. Very fancy stuff. And that took quite a lot of computing power and innovation and programming and development that was done that some of the teams I met were doing it in collaboration, where you had Ukrainian military officers and Ukrainian civilians. But also, you had programmers that were motivated just by their desire to help Ukraine or maybe they had some Ukrainian roots or they immigrated from Ukraine, say, living in Canada or the United States or Europe, and they were chipping in and coding and programming as they went to develop these more advanced systems. 

But the hardware, I'd say for this kind of stuff, is generally off the shelf. One kind of class of system that was widely available on the market and became very useful to the Ukrainians was agriculture drones. These are bigger things. For example, you can spray pesticides on your crops. So they already have this kind of function built in to release something from the bottom. Instead of pesticides, you're releasing - say, in one case, there was something called the dragon drone, which released burning fuel. Literally a stream of fire that it would shoot at enemy trenches. They were kind of tinkering with off-the-shelf technologies in many cases and adapting them to the battlefield. 

I'd say when you get into the more sophisticated battlefield awareness systems and command and control systems, there you really had programmers at top of the game working in different countries, different locations, and contributing to that effort to build really, how to put it, bespoke or unique systems specifically for the Ukrainian needs. 

[0:18:26] KB: Wow. Yeah. The Delta system sounds almost to me like turning this thing, this like real-life wide battlefield with all these unknowns, into almost like a video game setup, it feels like, where you could actually see this level of detail that to me always felt unrealistic in these games. But apparently that's real now. 

[0:18:47] SS: Yeah. Yeah. And the influence of video game culture and video game technology on all of these developments that we've been talking about is really amazing. I wrote an article recently for TIME magazine that I would encourage your listeners to check out. It's about the gamification of drone warfare. And it was a system invented by one of the key ministers in President Zelenskyy's government. He's the Minister of Digital Transformation, a gentleman named Mykhailo Fedorov. And I've gotten to know him very well. I've spent a lot of time with him. Young guy, younger than me for sure. What is he now? 34, 36? Anyway, young. Young by my standards. I don't know how young your listeners are. 

[0:19:26] KB: By mine too, but that's all right. 

[0:19:29] SS: He came up with this system of essentially assigning points to a confirmed kill. So, you had all these drone units that were working all across the battlefield. Some of them, many of them were sourcing their drones, their equipment from donors or buying them on the open market, tinkering with them, improving them. Some of them were buying them from manufacturers in Ukraine or abroad. But it was sort of a somewhat decentralized and dispersed effort. 

What Fedorov did is he first of all came up with a system for all of them to document and confirm their strikes. If they hit a Russian tank, there was a system that they would have to upload a video of that tank, that attack on the tank to confirm that it happened. And if the government was able to confirm that, they would give the team that carried out that strike a set of points. Say in the early days of the system, a Russian tank was worth 40 points. 

Okay, what do you need these points for? I mean, they're not just bragging rights. He then developed a system in parallel to this where these points could be used to buy - just like in a video game, to buy more drones, like Roblox or Fortnite, right? You can use the points you accumulate to buy or order additional equipment to then accumulate more points with more strikes. You see that? 

[0:20:56] KB: That's brilliant, though, because now you have an economy. The most effective strikers get the most drones. 

[0:21:02] SS: Yes, that's right. And there were a lot of debates. I talked to a few of the drone commanders involved in this. It had a lot of weird kind of unexpected, how to put it, side effects. One of which was that some of these drone commanders became kind of celebrities not only within the military but broadly in Ukrainian society. Because at the end of every month, the system would release, and it still does that today. At the end of every month, you get the list of the top 10 scoring drone units who have killed the most Russians or destroyed the most Russian equipment. And you have the name of that unit. And some of the units choose to stay quiet for security reasons. But I'd say at this point, at least half of them have decided to promote themselves as kind of warrior celebrities and use this also to raise funds from donors and say, "Look, we were number two on the rankings last month. We're the best. Give us more donations of various kinds," money, or equipment, or whatever from just regular people. 

That was an interesting evolution in the way the armed forces behaved. Usually, in a hot war, the soldiers aren't out there kind of promoting themselves. It seems like something from centuries ago with kind of this warrior class emerging. But that was one of the things that happened where now in Ukraine you have really a vibrant ecosystem of social media accounts, billboards, TV interviews of these drone commanders also positioning themselves as, "I'm the best. Look, I have the highest ranking in the game." Anyway, it's weird. But it's effective.

[0:22:41] KB: Well, it's fascinating thinking about all the different downstreams of this gamification. There's one. And I think I would bet, going further down, that probably helps improve internal morale within Ukraine and the folks there because they're like, "Oh, here's this celebrity. We're striking back. We're able to compete." All these different pieces there. 

[0:23:02] SS: Yeah. It gives society heroes to honor and to follow and to root for them and to support them financially if someone's able, both in Ukraine and outside. You're right. It is good for morale. 

[0:23:14] KB: Thinking about some of the other implications there, that promotes a culture of data collection as well. Now you have much - every individual is incented to make sure that they're keeping careful track of what is actually happening on the battlefield and reporting it up in a streamlined way. 

[0:23:35] SS: Exactly. And when I talked to the minister, Fedorov, who came up with the system, he said that's the most valuable thing for the government and the military because you see in real time what is working and what is not. Not only what unit is working, but what type of drone is having the most effect on the battlefield. And you're getting the most up-to-date data and the best documentation of each kill. 

As I understood from talking to these guys and some of the military commanders, there's often a lot of misreporting of successful strikes, generally in warfare, everywhere. Where in the fog of war, it's difficult to tell who really carried out some heroic strike, or was it just kind of hearsay or rumor, or there's no confirmation. This system really deals with that. And it tells you what is the most effective team. And then what the Ukrainians do is they do kind of sharing of expertise. 

Okay. For the last two, three months, this one team has been the most successful. They've come out at top of the rankings, and they've been using this particular type of drone. So then they do these kind of I want to say seminars or sharing of knowhow where these drone units then give instructions to others, less successful drone units, to try to bring them up to speed either by taking a member of another drone unit into their ranks for some period of time or having really actual classroom exercises where they're talking to each other or online sharing of information, some of which is classified, a lot of which is classified. But still, you have this kind of learning from each other, which is very good for effectiveness and has been a great help to the Ukrainian armed forces generally. 

[0:25:20] KB: To take this in another direction and geek out on something else. Another thing this reminds me of is a topic from machine learning. It's a reward function. It's a way that you can try a thing, see an outcome, and iteratively improve. Can you talk a little bit about the ways in which machine learning and AI are being applied in warfare in Ukraine? 

[0:25:42] SS: I'd say it's still pretty early days with AI. So, in this platform where drone units can use their accumulated points to buy new drones, part of the platform is classified, but it looks like basically Amazon. You can see some of it. It's called Brave1. It's called Brave1 Marketplace. And you can go on there and you can see what is available to the military units to buy. And there is a special section of AI enhancements. 

Basically, I'm not an expert in this, but chips and various equipment that you can use to enhance your drones and give them AI-enhanced targeting, for example, target selection and things like that. That has become available more widely to the Ukrainian armed forces to improve their drones with either computer vision, machine learning. All of that is being experimented with, but it has not been deployed at scale in Ukraine. There have been some reports that Russia is deploying it much more aggressively and in a centralized way. But those are - how to put it? The information on that is incomplete because Russia is so much more of a blackbox for journalists like me and for various intelligence agencies to try to follow this. 

But I know from the Ukrainian side, they're experimenting with it, and there are a lot of mixed opinions on how well it works. One thing I'll say is I visited a drone factory in Ukraine called Skyfall and they produce one of the most popular drones. It's got six rotors. It's a heavy bomber, so it can drop multiple explosive charges. The founder of that company told me, "Simon, I have AI targeting already." And he showed me the interface where there's basically a button. Not a button, but like a sliding scale where you can select how automated you want the targeting to be. 

For example, you can say, "Okay, computer, if you are 70% sure that the target you see in your camera is a Russian tank, if you're 70% certain, you can strike." Or you can say 30% certain. Or you can say, "No, you need to be 90% certain that that's a tank." And you can sort of slide it as you wish. He told me that he has not - and I checked in with him recently. This is still the case. He has decided on his own basically not to release that technology to his clients, to Ukrainian drone units, until the government or some other authority comes up with a legal framework for who is responsible for pulling that trigger. He doesn't want to be personally responsible for releasing the technology that allows, for example, a drone to make a mistake, to make a decision about targeting and firing a weapon, and then accidentally killing a civilian, takes the fault for that, or committing some other war crime or other atrocity. 

He needs some clarity on the legal side, and that is not yet there. But the technology in this way in many interesting ways has outpaced the legal frameworks that Ukraine is working within. I think other drone units and other drone manufacturers may have different views. But as you can tell, it's not centralized. It's sort of up to the individual commander, up to the individual manufacturer, to decide how much of this technology to unleash and how. 

[0:29:06] KB: Yeah, that is absolutely fascinating. It reminds me of a conversation I was having with someone in the topic of using AI to generate code and pushing that farther and farther and saying, "Okay, who's responsible? Right now, the answer is whoever says, "Yes, this is good code and commit it." But there's a future where this stuff is happening autonomously in the background and we have no framework for who is responsible ultimately. 

[0:29:30] SS: I don't see the legislative bodies of our planet coming up with those frameworks anytime soon. I don't see - I haven't heard about a concerted effort in Ukraine or elsewhere. The US military has some guidelines on this basically that a human needs to be in the loop on any decision to pull a trigger or drop an explosive or anything like that. But the Ukrainians don't really have that as explicitly. It's a little bit of a wild west situation there with AI. 

[0:29:58] KB: What about some of the other potential applications, things like autonomous navigation or that type of thing? 

[0:30:03] SS: One thing that I think some of your listeners would appreciate is in the context of this kind of gamification, I talked to one drone developer who had this idea that, "Look, Simon, we already have the technology to realize something that existed in a science fiction novel called Ender's Game." Have you ever heard of this one? It's also a film. 

[0:30:24] KB: I loved that game or that book as a child. I read the whole series. 

[0:30:28] SS: Yeah, there you go. So, for some of your listeners who may not be familiar, it imagines a futuristic world where there's a war between humans and space invaders, and the humans resort to essentially teaching children how to fly drone swarms that attack the alien invaders. And the children are made to believe that they're just playing a video game, that what they're seeing inside their controllers is not reality. It's just a simulation. Whereas, in fact, the game they're playing is actually manifesting in real life and they are controlling a real drone swarm out in space that's attacking the aliens. You get it? 

This drone developer in Ukraine told me, "Simon, we can do this now." You can be sitting in Kansas and controlling a drone or multiple drones from your headset or your computer and they're flying somewhere in the Ukrainian battlefield with a Starlink or some other communication system and you're just piloting them. And this drone developer suggested this could be a good idea because drone pilots are not as common as you'd think. Good ones, anyway. 

If you can tap into that pool of talent, he was saying, I'm not saying this is a good idea, but if you can tap into that global pool of talent, anyone who wants to participate in the war in Ukraine as a drone pilot can sort of maybe go through a test and apply. And then if you are a successful pilot, you sign a waiver. You are given command of a number of drones out in the front. And this developer said, "You might not know whether you're flying an actual drone or you're flying a simulation." 

I mean, that just blew my mind when I heard that. I was like, "Whoa, whoa, whoa." I had never read Ender's Game. I went back after this conversation and read the key bits and understood what he was talking about. But, I mean, man, it seems like too far on so many levels. But the technology is there. The reason we're not doing that or the reason Ukraine is not trying to do that is more legal, social and so on, right? The technology already permits that. 

[0:32:40] KB: That's wild. So in the Ender's Game, if I recall correctly, it was different in the sense that it wasn't drones, it was people on the ships. So in some ways, it's even more complicated ethically. You're commanding people to go and lay their lives on the line without knowing that they're people. In other ways, it flips it around because there is a train of responsibility. Those people are still making the decision. In this case, how does that even track? 

[0:33:02] SS: Yeah. Just to continue this thought experiment. And I'm not endorsing this. I can say for sure that it is something that people at a high-level in Ukraine in the military and in the military-industrial complex have been talking about. It frees the pilot from a sense of moral responsibility in a way. You're so detached at that point from the actual act of killing, the act of dropping a bomb on an enemy target, whether it's a person or a tank or a multiple rocket launcher, whatever. You're so detached, it's like playing a video game. 

And in some ways, I talked to one ethical expert who advises the Ukrainian military who said, "That's not a good thing because for various ethical and reasons related to international law." We do want the person responsible for pulling the trigger to understand the full moral weight of what he or she is doing. But in terms of kind of efficiency in avoiding things like post-traumatic stress and burnout, all those things that that many soldiers and drone pilots go through, there is the suggestion that maybe that's not such a bad idea. That debate is ongoing, is a live one. 

[0:34:13] KB: It's wild that we're in that world. 

[0:34:17] SS: It is. And it's coming faster, Kevin, than you - 

[0:34:20] KB: It's coming faster. Well, and playing out the analogy just even a little further. In the Ender's Game, that the crisis at the end was the lead character decided to destroy a civilian population because it was the strategic way to end things deterministically and he thought it was a game. 

[0:34:39] SS: I put this in the article, this conversation about Ender's Game. Just a paragraph very briefly. And my editor read that before we published, and she was like, "It's like the people talking about this never read the end of that book." Because it's not a happy ending. 

[0:34:54] KB: It's not a happy ending. 

[0:34:55] SS: It's quite grim. Yeah. What I find so disturbing and fascinating about reporting on this, the future of warfare, the way new technology is reshaping warfare, is just how fast things are evolving, especially in the cauldron of a war for survival like Ukraine is facing. It's facing an enemy that is genocidal in its intent and really wants to destroy Ukraine, wipe Ukraine off the face of the earth. 

And when you're faced with that, it really pushes you, A, to seek any means of defending yourself and to push aside legal and moral qualms that may otherwise be the subject of intense debate within the military or within society or within academia. Ukraine, to some extent, doesn't have time for that right now. And there are many people. I wouldn't say it's the majority. I've actually been surprised the extent to which they are very thoughtful about these decisions. I mentioned already the ethics adviser who's very well respected. And I quote him in my article, Gundus Mammadov. He has a strong voice and he's always there to remind the top brass in the military of the ethical implications. But there are cowboys for sure among the drone unit commanders on the ground even more so once they're in this kind of competition of the gamified - they call it the bonus program, the gamification of drone warfare, where they feel this competitive need to outdo their rivals, their other drone units in the field. And they might be pushed to use a technology like AI in a more aggressive way than others. 

[0:36:33] KB: I think there's a kind of interesting thread for us to pull on here around what that sort of process of innovation and feedback looks like. I think many of our listeners are familiar with innovation inside of something like a startup environment where you have at least financial pressure that reduces your timelines and forces you to focus. I hope sincerely not too many of the listeners, at least those who are not living in Ukraine, have had to deal with the level of pressure and adversity facing the folks doing this here. But I'm kind of curious, what does it end up looking like? What is the steps from, "I have an idea," to "I'm trying it," to "this is actually going all the way up to being propagated across the military command." 

[0:37:18] SS: Yeah. I mean, it starts often with - there's a startup ecosystem in Ukraine in the military tech world. And these are often people who start in their garages using whatever they can come up with, using their skills, calling their friends, getting donations, emptying out their savings accounts. But that's at the kind of really initial level. Before the seed funding. 

I will say that there have been basically accelerators or something you could compare to venture capitalists going around physically Ukraine. I don't want to call any of them out by name, but some of them are household names in the US. Some of them are famous three-letter agencies of the US government. And they were starting early in the invasion. We're talking like spring of 2022. Going around, looking for successful tinkerers, successful teams that were coming up with new solutions, new technologies. The team that first put a Starlink on top of agricultural drone is one good example. 

These people, let's call them venture capitalists. That's an imperfect term in this context, but I don't think there's another one yet that I'm familiar with. They are Westerners, foreigners. They come in, they look around, they take a tour of the front, and then they pick winners. They say, "Okay, these guys are doing something interesting." I have a conversation with the founder of this kind of startup team that's working on a new kind of drone. He seems like his head's in the right place. He's responsible. Let's give them a certain amount of money. And if they need some equipment, let's give them a bunch of chips, drones, whatever they need to accelerate their development. 

And I know that some of the names on the leaderboard of this gamification program, they did receive help from these kind of support systems or individuals. Sometimes these were nothing more unusual than like a crowdfunding website where you would have some people in the kind of charity or philanthropy community gathering money for drones or Ukrainian defense. And then they would say, "Okay, these guys in the field are doing something really interesting. Let's support them. Let's funnel the money to them." 

Often times it was also just much more military or intelligence services from the west coming in and saying, "These guys are doing really great work. Let's make sure that they don't fall behind." That is also part of the process that you've asked about of who's iterating, who's pulling ahead, and how. It's merit-based. I mean, it's not like someone's calling their cousin in the CIA and saying like, "Can you please help me out with my drone startup in Ukraine?" I mean, these decisions are, from what I've heard certainly, based on how successful is the technology that they've come up with. 

[0:40:06] KB: Yeah. Well, and you have once again a downstream impact of the gamification is you have the feedback loop. You have, "We're trying this. And the team's using this. Their score is going up. Or we're using it in our team and our score is going up." Wow. 

[0:40:21] SS: Yeah. And they see it on the platform, this Brave1 marketplace. A new technology comes up and they throw it up on the marketplace. Somebody tries it. They have success. They spread the word on social media like, "Wow, we decided to use a few of our bonus points to acquire this new gadget or this new kind of AI-enhanced targeting system and we just went up five spaces on the leaderboard." And then everybody wants to try that. It gives you the kinds of things that the US military takes years to figure out. 

The Pentagon will do studies and hire all kinds of contractors and bring in all kinds of think tanks to analyze. And then once they have an idea of what they need, they put it, they give a contract to Lockheed Martin or one of these big companies in the military-industrial complex. This stuff takes years to do the study, to figure out what works, to get the equipment, to get the approval for the budgets. Anyway, this is incredibly cumbersome, slow. By the time the weapon is then delivered to the troops, technology has progressed so much that new weapon may no longer be relevant. 

And just one more tidbit that really blew my mind here in this context, a lot of the - especially in the early days of the invasion, the US military and other militaries, the British were donating some of their drone equipment and other kind of more high-tech weapons, and a lot of them would just end up gathering dust in warehouses because they were not capable of overcoming the electronic warfare and jamming environment that Russia was throwing over the battlefield. These things, very expensive systems, loitering munitions and all these kind of fairly advanced drones would just not be useful. It wouldn't fly. The Ukrainians either had to re-engineer them to make them overcome those electronic warfare challenges. But one guy told me that it's often just faster and easier to build the thing from scratch rather than re-engineer what the Americans sent us in 2022. 

[0:42:27] KB: It's amazing to me the parallels between what it feels like, for example, building inside of a startup in Silicon Valley versus in a big company somewhere else. 

[0:42:36] SS: Yeah. 

[0:42:38] KB: There's the pressure cooker environment. There's money that's available that's going purely based on merit, not based on connections, though there are still some connections. I'm sure there certainly are in Silicon Valley. There's the culture of advice-giving and sharing, best practices, and trying stuff. It's really wild. I'm curious now. We've talked mostly about what has happened to date. How do you see this continuing to evolve looking forward both within Ukraine and then spreading out across the world? 

[0:43:08] SS: Yeah, great question. And it's something I've really been looking at very closely as we approach the 4-year anniversary now of this invasion. I can't believe it's gone on for this long. Currently, Ukraine has basically a total ban on export of weapon and other military technology. That makes sense. You're in the middle of an existential war. You don't want your weapons manufacturers to be shipping stuff to other countries. They should stay inside Ukraine and serve the war effort. 

It is possible to get an exception for this, but it's very difficult, very rare. But I know from talking to some of the Ukrainian manufacturers that there is a lot of demand for what they're coming up with and they're raring to go. The manufacturers, including the one I visited, Skyfall, have built additional capacity into their factories, more than they expect the Ukrainian military to need. So that when that export ban is lifted, at some point when the war ends, when there's a ceasefire, we're not sure how that will happen, but at some point, Ukraine will allow its manufacturers to begin exporting this technology, this knowhow, and it will go global. Believe me.

A lot of what they've come up with maybe isn't something that the US military or the Israelis or the French would buy necessarily because they have their own stuff. And it's impressive in in many ways, maybe more impressive. But what the Ukrainians have gotten very good at is cheap, easy to use, easy to scale, easy to tinker with and adjust based on the changing dynamics on the battlefield. And that is really a valuable commodity for not only governments in smaller countries. But also, there could be militias, there could be various organizations, let's put it that way, non-governmental organizations that want this stuff, want these weapons. 

I wonder sometimes about the level of influence or control that Ukraine's European or American allies would have in limiting who Ukraine or Ukrainian manufacturers trade with and share this equipment with. That could really begin to redraw the map of the balance of military power around the world in some interesting and disturbing ways. 

Now, the same goes for Russia. No one's going to stop Russia from selling its equipment. It is already doing that with North Korea. It's doing it with Iran for sure. There's a lot of technological knowhow. Just a small side note here. Russia was so behind in drone technology in the beginning of the war that Putin, Vladimir Putin had to personally go to Iran to ask the Mullahs to sell him drones. And they did. Just a side note. 

No matter how this war ends, I think it's very reasonable to expect that the global market for weapons will be transformed by the flood of capacity that both Ukraine and Russia are developing for drone weapons, electronic warfare capabilities, AI-enhanced targeting, all the stuff we've been talking about. It's going to become available either on the black market or just at kind of military trade shows to the highest bidder. And that's scary to think about, especially if you look at terrorist threats or threats from various militias like the Houthis in Yemen and so on. The Houthis have used drones to block shipping routes in the Red Sea in recent years, but their drones are pretty primitive. Imagine if they do get access through the black market or otherwise maybe through Iran to more advanced systems. That's all coming down the pike pretty soon, I'm afraid, in the coming couple years at most. 

[0:46:59] KB: It is one of those sort of shifting moments where a lot of the sort of established lines of power start to break and fragment because now there's a - you don't need a million dollars to buy a rocket. You can buy a ton of drones for a lot less. 

[0:47:19] SS: That's right. This really, on a kind of philosophical level, has been the main lesson of everything we've been talking about. The ability of a small nation, under-resourced nation to use this technology to totally turn the tables on a much more powerful adversary. Probably the most famous example of that that some of your listeners might remember, you might remember in early June of this year, Ukraine smuggled a couple of truckloads of drones into Russia and then released them remotely. The pilots were sitting in Ukraine. They were flown into military air bases very far from Ukraine inside Russian territory and used to destroy Russian bombers, strategic bomber planes that are part of the Russian nuclear triad. Triad being land-based nuclear launchers, submarine nuclear launchers, these long-range bombers, right? That is the kind of holy trinity of a strategic military power like Russia or like the United States. 

And Ukraine was able to use a couple of truckloads of these cheap FPV drones to destroy - it's estimated about $10, $12 billion worth of these strategic bombers. And forget about their monetary value. Their strategic value is what's shocking in this attack that was so successful because they hit right at the pressure point of Russia's strategic nuclear deterrent and they did it using FPV drones. It's just insane. In terms of what you just said, the rebalancing of military power and how this can really just shake up the entire chessboard of great power competition, military might and everything related to that. 

[0:49:05] KB: The new key resource may be the number of nerds you've got. 

[0:49:07] SS: Yeah. How to tap them, how to get them involved and coding. Not to create a unicorn in whatever the consumer market, but more in military. And in the United States, we're seeing that. I mean, I don't know if you sense that. I mean, it's more a question for you in your community. Is it becoming kind of cool to work for military applications? Is that more of an acceptable kind of career path for a really talented coder? 

[0:49:34] KB: It's a great question. There's definitely a lot more sort of military startup things going on. Which, to your point, a lot of the traditional military-industrial complex is very slow moving, is very maybe behind the cutting edge in a lot of ways. Or if they're at the cutting edge, by the time they actually ship it, it's behind the cutting edge. There's a proliferation of security and military related technology industries. I think it really depends on which subculture you're in, right? 

Let's say Silicon Valley is an interesting example where a lot of the folks on the ground and the coders tend to be more leftist, many more pacifist or other kind of idealist. Most nerds. Not all nerds, but most nerds, we're pretty - I'll speak for myself. I'm pretty pacifist. I'm pretty strongly in favor of thou shalt not harm others around you if you can help it. On the other hand, I do think anyone who's experienced bullying, either personal or whatever, can empathize with Ukraine and what it feels like to have a superpower coming in and trying to destroy things. And I think seeing what's going on there, seeing some of the conflicts around the world, there is a sense of like, "Hey, maybe we need to actually invest in defending ourselves a little more." 

[0:50:56] SS: Yeah. From what I understand from a distance, and I do want to do more reporting about this in Silicon Valley, how is the culture there changing? Alex Karp, the CEO of Palantir, published a book recently that dives into a lot of this and argues for Silicon Valley technologists and inventors having a moral obligation to help the US military maintain its dominance around the world. It's an interesting argument. I'm not endorsing it, but it has driven a debate in Silicon Valley as I understand and broadly in the military-industrial complex about how much the partnership needs to strengthen between, say, the Pentagon and Silicon Valley. It is strengthening. I just don't know if from a grassroots level the talented young coders and developers coming out of Berkeley, or Stanford, or MIT feel a kind of drive or allure to go in that direction yet. I don't sense that. It could develop in that direction, too. 

[0:51:57] KB: I mean, I think the thing that this conversation has reiterated to me is we can put our heads in the sand and choose to ignore the changes that are happening in the technology of warfare, or we can get with the program. I personally do not want to be working on military applications. And I think that's a perfectly valid personal framework. And if my homeland was being invaded, you'd better bet I would be there and be working on those things. 

[0:52:28] SS: Yeah, it'd be great if there was a kind of legal framework before we get to that point of an existential war. I think a lot of problems can arise for humanity if those innovations and technological evolutions happen in the context of a war for survival without the kind of legal homework and the ethical homework having been done beforehand. I think it might be more comfortable or more compelling for young developers to go into that field if they knew, "Okay, these are the parameters within which we're going to work." Whereas right now, in the application of AI in warfare, the development in drones, it really is kind of - yeah, it's a wild west. Like I said before, there really are not the kinds of guardrails that you would expect. 

We have, for example, in the field of chemical weapons or nuclear weapons where there's a very well-defined decades-old legal international law architecture that limits how those can and should be used, those kinds of weapons. Here, nothing like that exists. That's kind of scary.

[END]