Welcome to Tech Stuff. I'm as Valoshan here with Kara Price.
Hi Kara, hi os.
I'm excited to share today's interview with Dexter Philkins, who's a longtime war correspondent and now a reporter at large for The New Yorker. He recently wrote a piece to answer a very big question, is the US ready for the next war? In search of that question, Dexter traveled all over the world, from Israel to Taiwan to of course, Ukraine, where he visited a drone factory. Here's Dexter on how Ukraine became ground zero for drone warfare.
The Ukrainians, on their und really they have invented basically a whole new way of warfare. They build really chieved drones that are really really accurate and really really and they build them by the thousands and the thousands and the thousands. So last year twenty twenty four, the Ukrainians built and deployed two million drones on the battlefield. The moment the Russians start to move in any kind of numbers at all, they walk right into a wall of drones. Yeah.
I've actually seen videos of these drones. It looks like a swarm.
It really does, and there are these insane stories about how Ukrainians are protecting themselves from drones. In fact, old fishing nets from the south of France are being shipped to Ukraine and kind of put over roads to basically prevent the suicide drones from hitting people and vehicles on the roads. There's a great story recently in The Guardian about how proud the fishermen of Provence are to be
helping Ukrainians out with their old nets. There's also another detail in the story which I found just absolutely mind blowing, which is that drone operators in Ukraine are given points for kills and for destruction of tanks and other milestones essentially, and the units that get the most points get more drones. It's basically a video game mechanic for distributing weapons of war.
But the story is not just about drones. It's also about how the wider future of warfare is being driven by technology.
The most important thing is not the jet at the bomber, the aircraft carry that costs billions of billions dollars, the most important components of software.
On the one hand, it's inside baseball, but on the other hand, it's absolutely fascinating. The whole paradigm of arms manufacturing is being turned on its head. In the past, the government would decide what it wanted and then put in an order. Now the private sector is taking a build it and they will come approach. And this simple fact explains the explosion of venture capital funding into defense tech that we talk about quite often on the show.
Yeah, you mentioned Ukraine, but you know another place my mind goes to in terms of the future of warfare is obviously Israel.
Yeah, that's right. Deta talks about that about how the IDEF is using AI to identify and locate people with potential links to Hamas based on a number of factors.
Based on their telephone calls, based on the location, based on the buildings they go into. And there's just enormous amounts of data, so much data that's all being crunched into these big computers that then essentially spit out target recommendations.
And what did the IDF do with this information?
They use it to decide who to kill. Dexter was told that after the AI analyzes the data, it makes a recommendation on who to target, and then a team of humans in the IDF tries to verify whether or not these targets are really connects to hamas and whether or not to make the kill decision. Dexter reports, however, this IDF team would get about one hundred recommendations from AI per day, and that sometimes the latency between the AI recommendation and the kill decision is less than a minute.
You know, at that point, is there like even a human in the loop?
I mean technically legally, legally yes, but practically less than a minute, not really. So we talk about that, and we also talk about Palme Lucky. He's the founder of Anderil, Silicon Valley's hottest, most darling defense tech startup, Unicorn, And we talk about how all these different places in the world come together, So how the battlefield of Ukraine and the warring Gaza are ultimately affecting how the US and China think about a potential conflict in Taiwan.
It's fair to say it would be catastrophic for the United States and for the kind of the interests of the Western world.
Before we get there, dextro Nite started off by talking about Dexter's visit to a Ukrainian drone factory.
So you know, I was blindfolded. I was taking this factory. They took the blindfold off. I was inside this you know. It was like a small, small like warehouse. There were a couple hundred people in there, most of them women, and they were they were making drones. But the drones, it's just a little thing. It's nothing. It's a farb and fib or frame. It's like a foot and a half by a foot and a half. It's just a square. It's got little propellers on it. I mean, it's nothing.
And then it's got a bomb rack and then a camera so that the person who's driving the drunk can kind of see where he's going. That's it. It costs five hundred dollars and so with not a lot of money, they can deploy thousands of these things. And that's what they're doing. And I should say the Russians are doing the same thing. And so it's this incredible kind of dogfight that tapping in eastern Ukraine. But the technology for
now appears to favorite the defense. And so it's largely because of drones and drone swarms and the tactics Ukrainians have developed that they've been able to hold the Russians off.
You talk about it largely favors the defense, but there's this extraordinary David and Goliath moment where some of the cheapest technology of warfare knocked out some of the most expensive thousands of miles behind the front lines.
That's right, that's right. And so for instance, the guy that I spoke to, he was like running like a trucking company like when the war started, and he's like, well, you know, the Prime Minister called and they said, hey, can you make drones? And that was that was three years ago. He never made a drone his life. And so this was a government project that I saw with private companies, and they've developed these anti ship drones. They're like these little toy boats. They're like six feet long,
no people, a little cameras. Some of them are just loaded with explosives. Others had like kind of torpedoes on them. But they have chased the entire Russian Black Sea fleet, some of it out of the Black Sea and the rest of it all the way to the eastern side of the Black Sea, where they've basically reopened what the Russians had blockaded, which is the sea lanes that allow Ukraine to export grain, which is you know, lifeblood. It's
all the TRD currency. That's the kind of things that are happening here, and it's an amazing thing to see up close.
The other moment in the piece that really caught my imagination was the fact that the drone operators have a kind of internal economy and the more points they accrue, the more drones they can get for their unit. I mean literally like a video game.
It does feel like a video game. The drones are are flown to their destination by a guy on a joystick, you know, manning a joystick. They have a screen, they have a laptop. Often they're not sure what their destination is. They're they're looking for targets or more typically, they'll have a surveillance drone finding targets and then directing the other drones to the target. So tank spotted go here. That's
essentially how the game works. There's drone teams all over the front line everywhere, and the best drone team, the team with the most kills get You get points, you know, like you get a number of points when you blow up tank and you can prove it. You get a number of points when you when you kill a Russian soldier when you kill a missile battery and those points out up. So the team with the most points, they don't get a prize, you know, or they don't get
a cash award, they get more drones. And so you can see how how well that works. So that the really killer teams, and there there's some just utterly lethal teams there. They were they were just getting as many drones as they could handle.
One of the sort of iconic images I've seen of the Ukrainian battlefield is this crisscrossed FIBROTI cable.
Yeah, it's crazy. But the biggest challenge to be overcome if you're trying to run a drone war is electronic jamming. And so you're sitting there with your joystick and you're flying your drone around and you're looking for a tank or you're looking for a platoon somebody, and somebody to crash your drone into. The Russians are really really good at jamming the radio signals and so that renders the
Ukrainian drones inoperable. When I picked up the five hundred dollars drone in the Factory and Western Ukraine, it had a kind of frequency horp so like if it goes up on one frequency, so the guy with the joystick is steering the drone and then if the Russians jam that, he quickly bounces to another frequency. If they jam that, it bounces to another where the long cables come in and they're fiber op the cables, so they're just you know, super super thin. You can't jam fiber op the cable,
and you can. You can find us on the web. You can find video footage. But there's there's drone teams, like Ukrainian drone teams racing down roads in eastern Ukraine that they've covered with netting the whole road.
Of course, there's a big connection between jamming drones and the race towards autonomous weapons. I want to come back to Palmer Lucky and full autonomy. But in between is Israel and Gaza. And some have suggested or questioned that Israel is already letting a I make kill decisions on the battlefield in Gaza. I think in your piece, the answer is AI is identifying targets, and sometimes the gap between identifying the target and making the kill decision is
extremely short in terms of time. But nonetheless there is always a human in the loop. I mean, what's the difference between a killer robot and a robot telling a human who to kill.
I mean, that's the heart of the matter. So in Israel, you get a recommendation from the AI, the mother computer, and it says it's basically a coordinate. It doesn't tell you who it is because I don't necessarily know who it is, and I think typically it gives you a kind of percentage of certainty. So if you fire at this coordinate on the map, the chances are eighty percent you're going to kill a bad guy, or it's seventy percent you're going to fifty percent you're going to kill
a bad guy. And that's essentially what they're acting on. So what the Israeli say is that it's always the human that kind of pulls the trigger. What we as the targeting officers is to try to verify the target
recommendation from the computer independently. But I think my great impression of talking to them was that perfection is kind of not really possible in more time, and so what you have is like in the heat of battle, when that little waypoint the coordinate on the map is moving and you just have a few seconds to decide whether or not to kill it or not. How much independent verification can you do? Do you really have time to send another drone over it to look at it, or
do you just kill it? That's the really hard part right there. That's what's happening in the room where they're making these targeting decisions.
Do the algorithms have a built in calculation of chollateral damage? Is that information that the officers have access to when they're making these decisions, or is that not even consideration.
Now this is where it gets really really I think morally tricky. As I understood it from my reporting, the computer will say fire here, here's your coordinate on a map. You have an eighty percent certainty, but you might kill eight civilians and that has to be included in the decision that you're making. You know, if I were a newspaper porter, I would have just written a separate story on this fact alone. I heard this from an IDF
targeting officer. He said, we were allowed in the beginning phases of the war, and I can't remember the exact number, but it was very, very high. We were allowed to kill a target that was recommended by the computer. If we believed that there was a possibility that we were going to kill as many as twenty civilians. Now, you know,
hit pause there, because that's pretty extraordinary. You're basically saying, I'm going to kill the one bad guy who's standing in a crowded market, and I think we might kill twenty people, but we're going to do it. And that's a human decision. I mean, that is in fact to what we're seeing in Gaza. If you want to know why and how there's so much distruction in Gaza and so many civilians dead, that's the reason. It's the rules of engagement decision. But they just dialed it way way back.
And I had a conversation with an American senior civilian in the Pentagon, Biden and who had been involved in a lot of the decisions in Gossip War, helping these rallies, and he said, if we had ever in the course of say the Iraq War, the Afghan War, if we had thought that there was a chance we were going to kill twenty civilians, that a decision like that would go to the President or it would go to the Secretary Offense. He said, the Israelis are doing this every day.
Sometimes more than once. So that gives you a sense of how I think, extraordinarily loose the rules of engagement have been for the Israelis. And I think it's not as much a technological kind of decision as it is a human decision. I think. Yeah.
I mean the ciche about you know, technology is that these are tools and we choose how to use them, right. I mean a soldiers sitting in a bunker making kill decisions with an eighty percent probability it's a quote unquote bad guy and you know, full civilians or six cividians, eight civilians being killed. I mean the technological element kind of creates this remove from the consequences of a decision almost.
Yeah, you can't. You can't hear the screams. But you know, you mentioned the Obama administration when they have the drone more in Pakistan. They made a lot of decisions like this, and they're very similar decisions. I had lots of conversations with people during the Obama administration about what they were trying to do in Pakistan, and they're trying to kill all kinda guys in Pakistan, but it was kind of
the same. They'd have a drone in the air watching a house and they'd say, well, we don't really know who's in the house, but there's guys in a pickup truck and they come and they all have guns, and they're going in and they're going out. And the guys with guns, like when I left the house, they went to another place. They went to a military camp. We don't know who they are, but we think they're probably bad guys. Are the civilians in the house. We don't know the answer to that, but I think it's a
pretty well documented. Some of those guys traumatized by what happened because you know they're killing people. It's totally bizarre. You've got a joystick and you're looking at a video screen and then you blow up a house it's got fifteen people in it, and then you get in your car and like you're gone and have dinner with your family.
It's like you're removed from those consequences, but like you can see how it's still I think haunting a lot of the people that have been involved after the break.
Could technology be the spot that creates World War three? Stay with us. Let's talk about full autonomy, i e. Killar robots making their own target and killed decisions. Is it a technological barrier to arriving there or is it a rules of engagement barrier? And if the latter, how on do you see that holding.
For Well, that's a good question. I think ultimately, as was explained to me by Palmer, Lucky and others, like, we can do this, we can, and we are preparing to go fully autonomous. That's what we are making these things to do, because the technology is taking us there.
One of the things he said to me, and we were standing in the showroom in the Andrel's office, so we're surrounded by the drones and the pilots airplanes and the in the in the anti drones and all the stuff that they sell, and he said, every vehicle in here, i'm a will be cut off from communication, from controlling our own weapons. But what it means is that these weapons have their own brains and they will have to
go and find their targets on their own. And you can see all the problems that could potentially arise from that, because you're really talking about kind of autonomous warfare. And so I had a conversation with I think he's now the president of and Brian Shimp and Brian sketched out the following scenario, which I thought was really interesting. He said, look, just imagine a one hundred miles square in the Pacific Ocean. Imagine somewhere inside of there, in that vast stretch of ocean,
there is a fleet of Chinese ships. What Andreil weapons are being designed to do, and what they will be able to do is we will fire our missiles at those ships, and they will, independently of us, with no human control, they will find the ships, they will target the ships, they will communicate with each other and decide who's going to take which ones and then go destroy them.
But ultimately those are human decisions. And so in the Pentagon there's a whole executive order that Biden put out and then I think Trump slightly revised that basically kind of allows for full autonomy as I understand it. Palmer Lucky was really clear about this, which was this is where it's going full on killer robe on it science
fiction movie. These things are fighting each other and they're killing people, and like, the Chinese are gonna have theirs and we're gonna have ours, and the humans are going to be basically taking a ringside seat.
What was your sense of Palm and Lucky and what's the scale of his ambition and where is he on achieving end.
Palmer Lucky is a remarkable guy, and he's deceptively modest about the whole thing. Like when you meet him, he's just you know, he's like a tech guy. He's young, he's wearing flip flops. He's got like a flowered shirt on, he's got she had shorts on. When I met him, he was eating from a bag and took He's revolutionizing the way that America is approaching its wars and approaching
to wars in the future. And and so the system that we've inherited and the system that we have now is hundreds of people in the Pentagon will decide what they need. And so they'll say, you know, we need a we need a jet fighter. And it has to
do all these things. It has to fly at supersonic speeds, and it needs to be able to die at this particular angle, and it has to have a range, and one thing after another, and this takes literally years for the Pentagon to devise a new weapon, whether it's an F thirty five fighter jet or a new Navy ship or a submarine. It takes them years. And so the paradigm that I think is finally being broken is that, you know, the tech world moves very very rapidly. That
I'm talking about our tech world in California. They move it super high speeds because if you don't move a high speeds to go out of business. Pentagon moved very slowly, and these two worlds never really came together. That's where Palmer Lucky comes in. So Palmer Lucky has figured out a way to bring these two worlds together. And he did that by by standing the Pentagon procurement process entirely
on its head. And so instead of you know, hundreds of people taking years and years to design a weapons system, Palmer went out there and just built the most advanced anti drone and the most advanced pilotless airplane. He just built the most sophisticated one that he could and he went to the Pentagon and said, this is for sale. We can do these, and we can make them. We can make a ton of these really fast, really cheaply.
But take it or leave it here it is. And so instead of an F three five fighter jet that costs two hundred million dollars, you have a drone like they cost fifty thousand dollars. And what Palmer again is kind of figured out is it's all about the software. So they just have these kind of very cheap shells, you know, i e. The drone itself, the missile carrying its explosives, but the only thing that really matters is software so that they can constantly make it better. The
drone missed its target, We'll fix the software. We'll tinker with the software. How do we make it more accurate? How do we make it better? The Pentagon says it wants us to do something differently, We'll just go in and tinker with the software. And they can do that really, really cheaply for nothing. And that's that's how the Palmer Lucky stood the whole Pentagon procurement process on its head.
I mean, both China and the US have looked extremely hard at what's been happening in Ukraine. How are the two countries responding from a technology point into view? And where does Taiwan say.
In all of this, Well, Taiwan's it's right in the middle between China and in the United States. You know, you could like look out a map at the Pacific and go like, who cares, you know, it's just this big island one hundred miles off the coast of China. Who cares? And like, it's a democracy. It's an ally, a very close ally. And it's also they make something like ninety five percent of the most sophisticated microchips in the world are made in Taiwan and made nowhere else.
The big rilla in chip manufacturing, the most sophisticated, the best is TSMC in Taiwan. And I remember more than a couple of timewhen Hees kind of jokingly said to me, you know, if the Chinese invade, the safest place and the whole country is going to be in the basement of the TSMC building something like. But to answer your question, if you got to stand way back, you take America on one side, has basically dominated the Western Pacific, and those are the most important sea lanes in the world.
Those sea lanes are absolutely crucial to the world economy in every conceivable way. And the Chinese are undergoing a rapid massive military build up of their navy, the air force, everything. They're challenging that they want to push the United States out of the Western Pacific you can see the dilemma that the Americans have. The US Navy, Air Force, everything, the entire armed forces are built on this principle which is completely antiquated now, which is we have very few
but very expensive and very sophisticated weapons platforms. I mean, just the other day in the Red Sea, an aircraft carrier had to make a sharp term because they were fighting against the Huthi's, you know, and two f thirty five's slid off the deck and went to the bottom of the ocean. There's two hundred million dollars, yeah, in like five minutes. So that's the Chinese, I think, to their credit, really, they have designed their entire army forces
to defeat the United States in the Western Pacific. That is what they want to do. They want to push the United States out of the Western Pacific, and they have spotted and identified our vulnerabilities, you know, very precisely. So they have designed their armed forces to sink aircraft carriers and to take out very very expensive American plans. They have designed their entire armed forces to do that.
And so the United States is kind of suddenly and way way too they'd woken up to that, and that's what's driving all this what we're talking about, I mean, the Pentagon, It is kind of I wouldn't describe it as a panic, but I think, well, I think the smart people are paniced. You know, the smart people are pretty.
Freaked out, and they're actively shipping drones to Taiwan.
Now, yes, Andrew's sending drones to Taiwan. So if you take Ukraine as a template Russia, they come in with these very expensive, very expensive platforms like tanks, like big radar installations cost millions and millions of dollars, and the Ukrainians are taking them out with five hundred dollars drones. This is what the United States wants to do in Taiwan.
I think Sam and Paparo, who's the head at the Indo Pacific Fleet in the Western Pacific to the American forces, has said, I want to turn Taiwan into a hellscape for the Chinese and to make them and he said this, to make them utterly miserable until we have enough time to get the rest of our military out there. What
does that mean. It's super simple. Flood Timelong with drunes and so if the Chinese try to say, invade the island and there's you know, there's a lot of evidence, a lot of evidence to suggest that the Chinese are
thinking very very carefully about doing that. You flood the air, you flood the sea, you flood the beaches with unmanned vehicles to basically sink an invasion force, And that, in a nutshell, is kind of what the plan is now that you know the Chinese, no doubt they're trying to counter it to and I think to me, the kind of the most likely scenario is that the Chinese won't invade Taiwan, but they'll blockade Taiwan. They'll just cut it off and they'll basically dare us to run the blockade.
So that's that's a whole other ball of wax.
But to your point earlier about the advantage being for the defender in today's world of military technology, and all out of tech from China would make Taiwan into the defender and therefore trigger the advantages of all these very cheap, unmanned aerial and water drones. But on the other hand, if a blockade, they essentially become the defender, and therefore the Americans become vunderbirth they try to run the blockade.
Yeah, exactly. It's a sort of it's everybody's spending a lot of time thinking about this. I went to a wargame. Actually, the Pentagon has these wargames, most of which you are classified. Yeah, yeah, well well they I went to a I went to an unclassified wargame at CSIS, the Center for a Strategic International Stuff. It's a think tanking Washington, and it was fascinating, but also it was like pretty disturbing the whole thing. And it was just a war game, but it was like,
oh my god. The scenario was a blockade. It wasn't an invasion. Chinese started stopping ships in the in the Straits of Taiwan. So there was a Chinese team and there was an American team, and everybody had essentially weapons that each country is believed to have. But it was fantastically bloody. I think it's safe to say tens of thousands of people died on each side. They stopped the
game after like the four go rounds. I wouldn't say it was a draw because like neither side had stopped, but they really really blooded each other, and I think that was That's the one thing we can be sure about is if the United States and China go to war, it will be it would be an utter and complete catastrophe. And this was no nuclear weapons were used because it had sort of decided at the beginning of the war
game no nukes. Sometimes that prohibition is not in the war games that they have, and sometimes nuclear weapons have been used. The Chinese did not attack the American mainland, even though they are quite capable of doing that. The United States attacked the Chinese mainline, filled thousands of Chinese, and I think what kind of sobered everyone in the room, and these are people in the room where you know they were pros. What sobered everybody was how fast things escalated.
You know, you started with this like kind of nothing event where the Chinese navy like stops a Taiwanese ship, and then it like stops an American one, and then there's some confusion, and then they shoot an American plane down, and then the Americans shoot a Chinese and then it's World War three like rapidly, very very quickly. It's just you could feel it in the room. It just took off.
And so this is what happens in war. You know, wars get out of control, and the wargame that I went to the war between the United States and China got out of control very quickly.
So I have to ask you, what is the future of wolfare that you saw and how afraid should we be?
Well, well, it's all pretty sobering, you know. It's the old Latin phrase devis possum parabellum. If you want peace, prepare for war. And I think if each side would say, in this case, the United States and China, we're both preparing for war. Hopefully the two cancel out. Everybody's really sober about it. Everybody's really careful, and because the prospect of going to war is so terrifying to each country, that war will be avoided. That's the best case scenarios.
That's that's the best we're going to do. If the terrence fails, that's what everybody's trying to avoid. And I think everybody knows this, including the Chinese. I mean, I
think the Chinese have been very careful in Taiwan. You know, they've been increasing the pressure, increasing the pressure, but they're careful, like they haven't done anything rash, and I think it's for the reasons that we're talking about, Like they've done more games too and they know how ugly these things are and how unpredictable they are, how rapidly they can get out of control.
What's the biggest question for you.
The kind of unanswered, unanswerable question, is can andrel and the Palmer Luckies of the United States deploy those these weapons that we're talking about rapidly enough to deter the Chinese from taking Taiwan. Because I think if the Western Pacific were to be kind of fall under Chinese domination, that would be a catastrophe too. So and I think the best way to prevent that is for these weapons to get out there as fast as they can go to deter the Chinese from doing that. And so that's
really the question. Can they get that stuff out there and ready quickly enough before the Chinese are ready to go? Because the Chinese have been very very clear, as clear as day like they want Taiwan. They've set twenty twenty seven as the year that they wanted to be in Chinese hands. They couldn't be more clear about what they want, and so I think it's fair to say the United States is racing to kind of prevent that. But my sense is that we'll work out the technological staff for
these guys. Will the Palmer Luckees of the world will figure out the technology, but can they do it in time?
Text to filchis thank you, thank you so much.
For having me.
That's it for this week for tech Stuff.
I'm Cara Price and I'm as Velocian. This episode was produced by Eliza Dennis, Tyler Hill and Melissa Slaughter. It was executive produced by Me, Kara Price, Julian Nutter, and Kate Osborne for Kneidoscope and Katria novelle Va iHeart Podcasts. Jack Insley makes this episode and Kyle Murdoch wrote off theme song.
On Friday, instead of our usual weekend tech episode, we will be airing part one of the podcast shell Game. This season, host Evan Ratliff founds a startup run by fake people. Listen to see how close we are to AI taking our jobs.
And please do rate and review the show wherever you listen and send us your ideas, your suggestions, and your feedback to tech Stuff Podcast at gmail dot com.
We love hearing from you.