The Pentagon’s growing alliance with Silicon Valley, where AI and tech innovation are redefining warfare and morality.
In this WhoWhatWhy podcast, national security journalist Andrew Cockburn exposes the little- known links between the Pentagon’s military-technology complex at the dark heart of Silicon Valley.
Drawing from his recent cover story for Harper’s, entitled “The Pentagon’s Silicon Valley Problem,” Cockburn delves into the military’s controversial alliance with tech behemoths.
This partnership, in addition to posing ethical quandaries, is fundamentally altering the landscape of modern combat through the advent of artificial intelligence and related innovations.
This collaboration ropes in some of Silicon Valley’s most contentious figures, such as Peter Thiel and Palmer Luckey, further complicating the moral implications of this union.
Cockburn untangles the intricate web linking the military to the technology sector, probing the implications of their partnership for global security, the rule of law, and ethical constraints on the use of force, both domestically and internationally.
Apple Podcasts
Google Podcasts
RSS
Full Text Transcript:
(As a service to our readers, we provide transcripts with our podcasts. We try to ensure that these transcripts do not include errors. However, due to a constraint of resources, we are not always able to proofread them as closely as we would like and hope that you will excuse any errors that slipped through.)
Jeff Schechtman: Welcome to the WhoWhatWhy podcast. I’m your host, Jeff Schechtman. If we were to distill the essence of today’s headlines into a rich reduction sauce, the primary ingredients would undoubtedly include war and the ethics of war, artificial intelligence, technology, money, politics, and most of all, Silicon Valley. These elements form the core of Andrew Cockburn’s latest investigative piece featured on the cover of Harper’s Magazine entitled “The Pentagon’s Silicon Valley Problem.”
It tells the compelling story of the intricate intersection of technology, defense, and the ethical quandaries reshaping warfare in our contemporary era and the role of the leaders of Silicon Valley. This even reaches into AI. As we grapple daily with the implications of AI on education, employment, and autonomous vehicles, we may be overlooking one of the most critical areas where AI’s influence is rapidly expanding, the shadowy domain of the military-industrial complex.
Here the integration of artificial intelligence into defense strategies and tactics represents not just a matter of technological advancement, but a profound ethical and strategic puzzle. Cockburn delves into the pressing questions surrounding the effectiveness, ethics, players, and potential consequences of this technological fusion. In essence, he asks, what are the broader ramifications for all of us from Silicon Valley’s increasingly intimate relationship with the Pentagon.
Andrew Cockburn stands as a formidable figure in journalism and currently serves as the Washington editor of Harper’s Magazine. His distinguished career spans several decades, during which he has authored impactful works and reported extensively on national security, defense policy, and the complex geopolitical forces that influence global stability. It is my pleasure to welcome Andrew Cockburn here to the WhoWhatWhy podcast. Andrew, thanks so much for joining us.
Andrew Cockburn: Oh, not at all. Good morning to you.
Jeff: Well, it is a delight to have you here. Certainly, the involvement of technology and Silicon Valley to some extent in the military is in itself nothing new. It goes back even to the Vietnam War, as you talk about in the piece, set that context first.
Andrew: Well, yes, Silicon Valley, to use the shorthand term for sort of digital tech, I guess, was really the creation of the defense complex, of the defense machine going back to the, well, the electronics industry in World War II maybe, but certainly the semiconductor revolution which was very much spawned by the Defense Department, by defense contracts. And then the internet itself was started life as, well, it’s precursor of the internet was ARPANET, which was a product of the Advanced Research Project Agency, a Pentagon offshoot.
And, going further down the line, Google Search, Bryn, and his partner, their initial, their crucial work on their search technology was funded while they were at Stanford by the DARPA, the Defense Advanced Research Projects Agency of Google Earth, again, a program called Keyhole, which was basically a CIA program. So they’ve never been poles apart. In fact, they’ve been very close together. But there was a time in the ‘70s and ‘80s with the dawn of the personal computer when there was a feeling that this was something different.
We were getting away from government-dominated, high-tech, militaristic high-tech, and the person Peter was liberating, which indeed he was. We had our own now information, accessible, universally accessible, and freer, that we thought for a while at least. But now, as I talk about in the piece, the tech, Silicon Valley, and the Pentagon are locked and never tighter on the seamless embrace.
Jeff: And AI is, of course, artificial intelligence, a big part of this. And while the public may think of AI, at least in the popular imagination beginning somewhere around November 22 with the launch of ChatGPT, the involvement of AI certainly goes back much further. And its involvement with the military goes back further.
Andrew: Oh, yes. Well, the military has longed for AI forever. They’ve had AI task forces going on for at least, well over a decade as far as I can remember. Yes, because the military, they’ve longed to pursue the dream of being able to control everything with a machine that somehow all information can be fused and interpreted for them to tell them what to do, which is, shows how uninterested they are in war really because war, the battlefield is all about menace and uncertainty and all that.
But anyway, and the example I quoted might be, I’d say a decade, I should say half a century, because during the Vietnam War story I love because it was the precursor to so many things during the Vietnam War, they concluded wrongly as it turned out that the Vietnamese fighting in the South absolutely depended on supplies coming down the Ho Chi Minh Trail from the North Vietnam through the jungles of Laos. Actually, most of them were coming through by sea through and then through Cambodia, but never mind.
So they said, well, what we have to do is to fence off that to interdict that flow of supplies. And to do that, we’ll set up an electronic fence. We’ll scatter, they did scatter thousands, tens of thousands of sensors across the jungle. Each of them was there to detect various signs of human activity like footsteps, the sound of feeling of traffic disturbing the ground, or the ignition spark of engines, or human urine, the ammonia smell from human urine.
And each of these sensors does a little radio transmitter which was beamed back to aircraft circling overhead which in turn transmitted that information to giant computers in a secret base in Thailand, the most powerful computers in existence, which were meant to analyze all this information coming from the different sensors and say, “Oh, we detect a Vietnamese column in such and such a place.” And then aircrafts, including drones would go and attack it.
Well, it was a complete bust. The Vietnamese figured out what was going on in a week, actually, and took steps to decoy it. They would run herds of cattle sort of ways away from the trail. So to fool the sensors that the troops were over there, they’d hang buckets of urine in remote areas, they would think they were there. It just didn’t work. It cost billions and billions of dollars and was a complete bust. But the military never gave up pursuing and investing in AI. Well, this is proto-AI, but the idea of computerized fusion of information until we get to today. I could take up all your time with the many failed projects and they’re still failing today.
Jeff: And bringing it up to the present in terms of failures, the Israelis were using AI as a defense protection on October 7th. That didn’t work very well either.
Andrew: No, they did. They tried this a few months before. They had Shin Bet, their internal security intelligence service, proudly announce that we’ve now got our own AI, a version of ChatGPT, and it can tell us exactly what the terrorists, meaning Palestinians, are doing, where they are, every motion and action they take. And that’ll tell us what he’s going to do. That’ll enable us to predict the future. And, of course, it didn’t because it’s all absurd.
They obviously fall into the notion that like the military here and like so many other people now talking about AI, that AI is a thinking machine. It’s not. It’s just on a vast scale. It assembles information and sorts it and comes out with a probable conclusion, but it’s not thinking. And so in the case of the Israelis, not just AI they ignored. AI was giving them the wrong conclusion that Hamas would not attack, and they ignored what they did ignore was all these very obvious signs that Hamas would attack, including Hamas’ own videos which they were putting online showing them practicing breaking through the fence and attacking settlements and taking hostages and all the rest of it. So AI just got them nowhere.
Jeff: Yes. In spite of these failures, there is a large contingent, both in the Pentagon, and has been for quite a while now, a contingent there and a contingent in Silicon Valley, a group of people that are pushing for a closer and closer relationship between Silicon Valley technology and the defense industry.
Andrew: Well, what I think is, yes, that is what you say is true, and certainly all the big tech firms are very eager and have been increasingly certainly over the last decade to start leaching off the taxpayer big time. And I think what’s going on, there’s a couple of things. One is commercial tech. I mean, all the sort of wonderful gizmos they brought up the smartphone, particularly all the applications and so forth. It’s been argued to me that the pace of innovation has really slowed down in the tech industry. That’s not completely true, it’s not across the board, but to a large extent, it is.
And so in that case, they look around and they think, “Well, if we’re not going to be able to come out with a new whizzbang thing to sell,” not like say the smartphone has changed everything, “what’s a really dependable source of revenue going far out in the future?” Well, the government, the taxpayer, you and me. So, they’ve been very eager to — they’ve all started to go to defense business and most particularly the cloud business. They’re all bidding.
There was a very huge bitter fight for the Pentagon Cloud a few years ago, particularly between Amazon and Microsoft. Microsoft got the contract, a huge $10 billion contract, and then Amazon protested saying Trump had unfairly tipped the scales because he didn’t like Amazon. So eventually the Pentagon realized it couldn’t afford or didn’t want to antagonize the losing big tech firms, so they’ve now parceled it out and everyone gets the slice of the action. It’s called a very unwieldy acronym, the Combined Joint All Domain Monitor and Control Network.
And the idea is way back to the Vietnamese defense, which was called Igloo White. This is great, great grandson of Igloo White. It’s the idea that all information can be fused and everyone participating in the defense efforts in a war can have access to everything that’s going on. Which it would take another show to explain why that’s an incredibly stupid idea too. So what we have therefore is now a big lobby.
Tech has become part of the military-industrial complex, and there are various unsavory signs of that, like the eagerness now of retiring senior officers, three and four-star general officers and below, to make their way west to California and get signed on with a venture capital firm or a tech startup, the revolving door, so they can sell their contacts and access and they can lobby therefore and help themselves, so they hope a huge payday, because they want this idea that you work in tech, you become a billionaire. So, you’ve got a lot of greedy retired generals and admirals who are hopping on the gravy train.
And the ideas that are coming up, most of them are perfectly stupid. The sort of ideas that I just read in The Wall Street Journal a couple of days ago that the venture capital is pouring money into hypersonic research. They’re going to produce hypersonic weapons. Well, first of all, research into hypersonic weapons is incredibly expensive, because you need all kinds of specialized wind tunnels and so forth. And secondly, no one’s ever managed to make it work. They say the Chinese and the Russians are way ahead of us, but when you look closely, certainly the Russian one, what the Russians claim is a hypersonic weapon, the Kinzhal being used in Ukraine, it’s not that.
I mean, it just goes very fast, but the whole idea of hypersonic, without boring you with the details, is something that goes super-fast and maneuvers, so it can’t be shot down. Well, the Russian thing, which is actually an aged missile, sort of upgraded and renamed doesn’t do that. And it’s unlikely there’s been any years now failed tests by Lockheed, for example, trying to get one to work. And they’ve failed.
So how some bright-eyed young entrepreneurs from Silicon Valley are going to accomplish it, I don’t know, but they’ll certainly get a lot of money out of the Pentagon.
Jeff: And talk about the push for this within the Pentagon. There was a lot of effort in bridging this gap between Silicon Valley and the Pentagon during the Obama administration with high-ranking Defense Department officials like Michèle Flournoy and Robert Work. Talk about that.
Andrew: Yes. Well, within the Pentagon, there’s been for forever, but more and more these days, a love affair with high technology. They love high tech because, well, really there’s money in it. They may not be directly or indirectly profiting, but the whole impetus of the Pentagon has been to develop whatever more complex weapons, which means the more contracts for a whole bunch of people, which invariably are over ambitious of how they sell them. You know, “We’re going to have a self-stealth plane that can also be a fighter plane that’s also a nuclear bomber,” to name but one concept.
And the trouble is that these things always cost more than projected. So you end up with less than you — fewer things, whether planes, ships, whatever, than you issue or project. But instead of saying, “Well, maybe we should do it another way and go for simpler concepts.” They say, “No, we need even more high complex things.” So, therefore, it’s not surprising. And the main spring of this line of thinking was an office in the Pentagon called the Office of Net Assessment, which was run by a guy called Andrew Marshall for years and years and years. And he would always come up with grand intellectual concepts, which as it so happened, gave a justification to the military for buying more complex weapons.
Like, “Chinese are ahead of us in precision weapons. We need something even fancier to do the [unintelligible 00:17:06].” And people like Robert Work came out of that stable. And Flournoy, she didn’t actually work for Marshall, but that’s the abiding doctrine in the Pentagon. If you want to get ahead, that’s the way you have to talk, because that ultimately is where the money is. And the people who — the industry, the whole complex, the military-industrial complex, exist to generate more money for itself. That’s its purpose in life. And so, these people, the people you mentioned, Work or Flournoy, there are many others, rode on that.
And therefore, it’s natural that they would sing the song to Silicon Valley innovation and the wonders of tech, and they can solve all our problems. And by the way, get themselves as advisory, consultancy contracts along the way.
Jeff: And talk about the other side, the push on the Silicon Valley side with people like Peter Thiel and Palmer Luckey.
Andrew: Peter Thiel, he made his initial money out of what, PayPal, I think.
Jeff: Yes.
Andrew: And then he had this idea that he wanted to be part of the military-industrial complex. So, he set up Palantir, in what, 2003 or ‘04, I can’t remember, around then, which was basically a data faulting firm to use. It came out of a technology that the fraud detection technology they developed for PayPal. And it could analyze patterns, which is what AI does, the essence of AI. And so, he’s always been doing that. Never actually made much money for a long time.
But he then later on, in what, 2013, 2014, along came Palmer Luckey, who’s kind of a prodigy. He developed the Oculus virtual reality headset when he was, I don’t know, 17 or 18, and sold to Facebook, to Meta for, I don’t know, several billion dollars. And then decided that he really, by gosh, knew a lot about defense, and knew what needed to be fixed, and knew what he could do with all his technological brilliance. And so, with Thiel’s backing, he set up this company Anduril. And he’s obviously a very smart fellow, I suppose, but the defense ideas or defense technology he talks, he sounds like absolute nonsense, he wants to…
I quoted the piece is, his dream of a superhero soldier, the soldier of the future, who will be equipped with the technology to know absolutely everything that’s going on around him, and where the enemy is exactly which hill the enemy is hiding behind and what his commander wants him to do, and what he should do next. Which I’m afraid, Brown practical expression in a defense program, blessed by the Pentagon, which actually went to Microsoft called IVAS, the Integrated Visual Augmentation System, which is a super goggles, plus a battery, plus a computer that the parole infantryman has to lug around.
And which is supposedly going to give him all the information he needs about the battlefield, and how much ammunition he has, and where his other comrades are, and where the enemy is, and what tree the enemy is hiding behind, and what his next mission is. Meanwhile, the poor fellows trying not to get shot and trying to find a rock he can hide behind, so. And the soldiers have been tested on, have complained bitterly, but that doesn’t seem to make much difference. And they’re still going to have to carry around a 50-pound battery pack. And there’s $23 billion of our money going into it.
So in the end, I see this whole fusion of eruption of Silicon Valley into the defense complexes, meaning nothing good for defense or certainly the poor soldier who’s going to have to use it.
Jeff: And one of the other things that we’re seeing is a pushback from time to time in Silicon Valley about the ethical issues surrounding this. We saw it several years ago at Google, and there are flareups from time to time. Talk about that.
Andrew: Yes. Going back to what I was saying earlier about, there was this dream for a while that tech was different, and was liberating, and was the antithesis of Big Brother, and the whole government machine, and people like Apple and they all said, “We never want government contracts, dealing with the government and horrible neuro hopeless solace bureaucracy. And we much prefer to operate light-footedly on our own.” And that feeling persisted in a lot of the workforce, people who wanted to work in tech.
So when it emerged that Google was secretly working on a — got a contract for something called Project Maven, which was going to use the wonders, basically the wonders of search technology to be able to analyze video footage, data, but initially video footage taken by drone, prospective enemy areas, and would be able to analyze it without human intervention to say, “Aha, the enemy tank is there,” or “The enemy is doing something else over there.” And that was Project Maven. And when that leaked out, we just had a secret select team working on it, when that leaked out to the broader workforce, there was an explosion of protests.
People said, “The motto of Google was meant to be do no evil. This is evil. We don’t want to work on weapons or weapons technology.” And 4,000 people signed a protest and a few, actually a very small few, still a few actually quit. And Google apparently said, “Okay, we’re sorry. We’re sorry. Okay, we’ve changed our minds. We’ll give up that contract, or least, we won’t renew it.” And a couple of years go by and then quietly Google bids on this thing I mentioned earlier, the joint all domain commander control. So, they were just as eager particularly Pentagon Cloud contracts as anyone else.
Google feels that they’ve fallen behind Amazon and Microsoft in that business, and they’re eager to catch up. So, yes. And I quote from a former Google person who quit just because he was so appalled at Google’s behavior. Actually, he quit, particularly because Google, while — to digress for a second, a lot of this is tied up with traditional threat inflation, which is all part of the military-industrial complex game, which is, “Oh, the enemy’s drawing ahead of us.”
And so particularly the whole people pushing defense tech talk a lot about China, and how the Chinese they’re so smart, and they’re working so hard, and they’re drawing ahead of us in AI and everything else, and we’ve got to buckle too and work harder, and spend more money, and catch up. And that was particularly true from the guy who had been head of Google, Eric Schmidt, and he is always talking about the Chinese threat.
Well, then it turned out that meanwhile, they were working on a Chinese contract, a contract for the Chinese to develop a version of the search engine for China, where you couldn’t look up politically awkward things like student protest, or freedom of speech, or anything else the Chinese authorities didn’t like. It was a thing called Project Dragonfly. So the guy quotes someone who went over that. But it’s just the hypocrisy that runs through the industry, it turns out.
Jeff: Bringing it back to the early days of this, one thing that seems to be different is that in the early days of the relationship between technology and the Pentagon, when you talked about the Vietnam War era, and there were companies like HP, and Raytheon, and all these companies that were engaged with the Pentagon and ARPANET, that in many ways what the Pentagon was doing and the government money, was driving innovation in technology. You don’t get that sense today. There’s much more of a sense that Silicon Valley is simply trying to sell their wares to the Pentagon, and that it is not a case of the Pentagon’s money driving any kind of innovation.
Andrew: Well, yes, that’s a good point. I mean, they might argue that by the Pentagon, by pumping money into AI or these other things, that they are in a way driving innovation, but I agree with you. They don’t. It’s like the great wave of innovation that’s swept through American industry, particularly in technology during the Cold War, has really died away, and all we’re doing is throwing more money at it. It’s interesting, why did this area of Northern California, basically the peninsula south of San Francisco become the heart of this?
Well, it was a fusion of — there was already big defense interests there. Lockheed had plants there, for example. You had Stanford, the whole intellectual, mighty intellectual engine of Stanford University, into which a huge amount of government research money was being pumped. And you had, the people wanted to work there. You had the counterculture, people that have a very innovative state of mind. I think that so many things came together. And they always try to reproduce that. There’s a huge effort around Austin, Texas, because that’s a tech center and that’s where they put the Army futures command, which is yet another government effort to sort of leach up tech ideas.
Now, this is no one center really, it’s spreading around the country. But I think I agree with your point.
Jeff: And finally, AI seems to be becoming a larger and larger part of this. In your conversations, in your research on this piece, talk about the fantasies about AI and what the government hopes to do with it at this point.
Andrew: Well, what they had to do, they hope to sort of — I quote a long pin of praise to AI by Michèle Flournoy who’s basically been a sort of — well, she is an arm salesman actually. She’s part of a thing called West Exec, which is basically a bunch of arms traders. But she says, “AI is going to able to enable us to do our intelligence. It’s going to be able to enable us to design planes better. It’s going to be able to enable us to maintain our weapons better, plan production.” Which is all nonsense.
And they have, they installed — well actually three years ago now, they switched on a big AI program in the Pentagon called Game Changer, which was meant to be able to tell them how the Pentagon works. Like the thousands and thousands of directives on what that policy on anything particular circumstance might be. And so, this Game Changer, may rather than sort through a mile-high stack of documents to find out what that policy should be about something, they can ask Game Changer, which will hopefully give them the right answer.
But one of the things that Game Changer is meant to do is to tell them where their money goes and what they actually spend all their money on, because they don’t know, because for years, they’ve been failing their audit, even though Congress has passed a law saying they have to be able to pass and award it. And so they installed Game Changer in 2019, which is going to tell them where the money goes, and they’re still failing their audit. So, it can’t even do that. And to top it all the press release announcing Game Changer, they couldn’t even — they said it was all very — what did they say? They’re using authoritative sources and they managed to spell authoritative wrong. So much for AI.
Jeff: And do you think any of this technology is going to change the future of warfare very much?
Andrew: Well, the future warfare changes all the time. I mean, yes. I mean, I said imagine though I’ve been blindly condemning the whole thing, there’s ways it can help with pattern recognition if you know what patterns to ask for. I mean, it’s the old rule of garbage in, garbage out. So, if it’s properly guided, yes, it can help. If you look at Ukraine, what has Ukraine taught us about the future of war? Well, you hear a lot about — first of all, you hear that, oh, you should hear that, what’s really been the determining factor in Ukraine is artillery and how much either side has of it. And artillery has been around for quite a long time.
The targeting has improved. So now, artillery can be very precise, but so therefore people are having to learn to camouflage themselves. Another very old technology. The other thing that’s made a big difference or seems very innovative is drones. But not the super high-tech drones that we’ve been getting all the publicity in the last 20 years, but cheap mass-produced drones like you and I can buy on Amazon. And they’re the important ones on the front line because individual soldiers can use their drone to see, to scan the battlefront and see where the enemy is.
And if their drone is armed, crash the drone into the enemy tank or trench or whatever. So that’s been important. But whether, I mean AI, I mean the technology in the end, yes, it’ll make some kind of difference, but I can guarantee you not in a way that’s being projected by Michèle Flournoy.
Jeff: And we should point out for our listeners that Michèle Flournoy was Under Secretary of Defense in the Obama administration.
Andrew: Well, that’s right. And she almost became the Secretary of Defense in the Biden administration till people pointed out her multiple conflicts of interest. And so they had to withdraw her name, but yes, she’s very much at the center of the order.
Jeff: Andrew Cockburn, his story in the current issue of Harper’s is “The Pentagon’s Silicon Valley Problem.” Andrew, it is always a pleasure. I thank you so much for spending time with us here on the WhoWhatWhy podcast.
Andrew: You’re more than welcome. Thank you, Jeff.
Jeff: Thank you. And thank you for listening and joining us here on the WhoWhatWhy podcast. I hope you join us next week for another WhoWhatWhy podcast. I’m Jeff Schechtman. If you like this podcast, please feel free to share and help others find it by rating and reviewing it on iTunes. You can also support this podcast and all the work we do by going to whowhatwhy.org/donate.
