Anduril’s Luckey on His New Company, Defense Innovation


Palmer Luckey, the founder of Anduril Industries who also founded Oculus Rift VR — which he sold to Facebook — discusses his new company, defense innovation, the Defense Innovation Unit and more with Defense & Aerospace Report Editor Vago Muradian at the Reagan National Defense Forum at the Ronald Reagan Presidential Library in Simi Valley, Calif. Our coverage was sponsored by L3 Technologies and Leonardo DRS.

Palmer Luckey

Anduril Industries

Reagan National Defense Forum

November 2018

Vago Muradian:  Welcome to the Defense and Aerospace Report.  I’m Vago Muradian here at the Reagan National Defense Forum at the Ronald Reagan Presidential Library in Simi Valley, California, America’s premier gathering of defense leaders, military, civilian, lawmakers, industry technologists, thought leaders, and we’re talking to one of those technologists now.  And by the way, our coverage here is sponsored by L3 Technologies and Leonardo DRS.

And one of those technologists and innovators is Palmer Luckey who was the founder and the driving force behind Oculus Rift, a game-changing technology and now is Anduril which is an innovative company that you’ve started since which has a little bit of DIUx influence in it.  And for the people who don’t know where that name comes from, where does that name come from, Palmer?

Palmer Luckey:  Anduril is Aragorn’s sword from the Lord of the Rings.  It translates in Elven to Defender of the West or Flame of the West.

Mr. Muradian:  I only wish I knew more Elven, and we won’t get into my knowledge of Klingon, not that I would confirm that or not, dating me thoroughly at this point.

Talk to us a little bit about the nature of innovation.  Right?  I mean you started a very, very innovative company.  You’ve now moved to start another innovative company.  And this is time when in a great power competition mode everybody is talking about harnessing innovation.  The last administration talked a lot, set up organizations, set up processes to try to capture it, DIUx being one of them and I know you worked closely with Raj Shah who we talked to who is now doing Arceo, his new artificial intelligence company. Talk to us about the nature of innovation and how the government needs to think, and especially the Pentagon, needs to think about innovation.  Because when you look at it on a commercial cycle, it’s operating at a breakneck pace. Whereas when you get it into DoD, not so much.

Mr. Luckey:  Innovation isn’t just about doing the right thing or doing the smart thing. A lot of it is actually about the business incentives and the structures.  I think right now a lot of defense programs are done on cost-plus contracts that incentivize very long schedules, building custom parts whatever you can justify rather than what you need to.  Doing excessive testing whenever you can justify it, rather than when you need to.

If you compare the results of those types of programs with the results of the consumer technology industry, the results are striking.  I mean at Oculus we weren’t trying to build something to maximize the contract value so that we could get our government allowed eight percent margin on top.  We were trying to move as fast as we could on our own schedules with our own money, build the best thing as fast as we could so that we could get it out to people. That’s how you build really good technology.

Imagine if the iPhone had been built on a cost-plus contract that was dictated by the government with milestones and deliverables dictated by the government, it just wouldn’t have worked.  And so I think that the government can learn not just from the superficial techniques. You know, the things that make Silicon Valley special, it’s not the smoothie bars and the massage parlors and the free lunches.  That’s not what drives innovation.  What drives innovation are incentives that reward employees who move fast, who do smart things and are all moving towards a common goal.  So that’s one of the things that I think you can learn.

And at Oculus we did a great job.  I sold Oculus to Facebook for a few billion dollars back in 2014.  I started Anduril because I saw these traditional defense primes really doing a good job of building iterative improvements, or I guess an okay job of doing iterative improvements on things that already existed.  You know, faster fighter jets, deeper submarines, bigger aircraft carriers.  But they didn’t have the technology and the tools and the talent that they needed to build systems around autonomous weapons and artificial intelligence and sensor fusion, and those things are really important parts of the future of defense. And the people who would be good at building those things are all stuck in Silicon Valley at companies that don’t want to work on national security.

So I wanted to take a place where we could take those Silicon Valley-style incentives and structures and bring them to the national security space and solve these problems faster and better than anyone else.

Mr. Muradian:  And what are some of those technologies and approaches and ideas you want to bring to market as you start the new enterprise which is looking at solving, scratching national security itches.

Mr. Luckey:  One of our biggest focuses is making people work with autonomous machines better. Letting machines do what machines do best; people do what people do best.

So our core product is something called Lattice which is an AI-powered sensor fusion platform that can take date from thousands of sensors and merge them into a single cohesive, real-time, 3D model of large areas, and then tag everything in that model with metadata so that you know where all the people are, where all the cars are, where all the drones are, and you’re able to predict not just what they’re doing right now but what they’re going to be doing in the near future.  And then push that data to mobile devices, to command and control centers, and to augmented reality heads-up displays on ground troops so they know exactly where the threat is, where their allies, are where the partner force is, where the dangerous places are, where the safe places are.

We want to build things that allow people to do what people do best, and machines do what machines do best.  I think that is going to be the crux of future defense technology.

Mr. Muradian:  How do you, Vladimir Putin has talked about artificial intelligence being a key national security —

Mr. Luckey:  I’ll interrupt you.  I remember the exact quote.  He said, “The leader in this sphere will become the ruler of the world.”  That’s James Bond villain type of stuff, and he’s saying that not because he thinks that he’s saying it from a position of weakness. He believes that he’s in a position of strength.

Vladimir Putin would never say that if he thought the U.S. was going to beat him at AI. And we should be terrified of that, that he is so confident in their ability to move faster than us on autonomous systems that he’s saying effectively they’re going to rule the world by beating us on AI.  That’s crazy.

Mr. Muradian:  Well, Rafal Rohozinski, we talked to him when we were in Canada, and one of the things he said is look, the irony of some of these sanctions is the Russian cyber, AI, everything is accelerated because they don’t have access to this.  [He] has also put in a paramount order —  he’s a paramount leader, right?  He’s put in and said hey, 20 percent of the national investment is going to go in AI.  So as you’re somebody who’s working in this space, how do you stack where the United States stands vis-à-vis its potential adversaries in what is a game-changing technology?  We were talking about man/machine interface.  There were friends of mine at a senior level who were talking about man/machine interface while at the same time being terrified at how adversaries would use man/machine interface.

Mr. Luckey:  Sure.

Mr. Muradian:  So talk to us not only about where adversaries stand but what are the things the United States clearly has to do to make sure it stays in the lead in this?

Mr. Luckey:  There’s a few things.  First of all, we need to set the norms about ethical use of autonomous systems.  We need to figure out how we can, if not having a person making every decision, at least have every decision that a machine takes go to a human at the responsibility level.  A human needs to be involved; a human needs to always be responsible.  You can’t ever say oh, the machine screwed up, nobody’s responsible.  You can’t have infinite diffusion of responsibility because that’s how you get to really scary ethical outcomes.

The other thing that we need to do is figure out how we can apply U.S. advantages that we already have to artificial intelligence.  China has a really big advantage that we don’t have.  With artificial intelligence the more training data that you can gather the faster you can move on certain things.  And they’re already using AI for, starting to do mass surveillance of their population.  Learning how populations move, how people move, how people interact.  They’re going to have a massive training set of over a billion people and we’re not going to have that in the U.S. because we are not willing to do mass surveillance of our own citizens.  China has no such qualms.

So if we’re going to compete with them, we need to start applying AI right now in the areas that we do have a lead.  In industry, in enterprise, in national security.  We need to start using our current military advantage to train future AI developments, and we need to start learning.

The U.S. is unprecedented in its ability to project power around the world and to participate in multiple large-scale conflicts at once.  China hasn’t had to do that for a few thousand years.  Russia has a little more recent experience, but I think we can keep that under control.  We need to leverage the advantage that we have into artificial intelligence as quickly as possible.

Mr. Muradian:  What does the world look like in ten years?  We’re talking about commercial cycles.  There’s Moore’s Law, but it’s really extraordinary the pace of change, right? Just looking at my career as a reporter, the amount of change, you know, we used to like have to carry pocket knives and have little alligator clips to send our stories in on our [Trash] 200s or [Trash] 100s because they were more portable in order to be able to feed stories. Now we’ve got super computers in our pockets with global connectivity and you’re really not out of reach in any way.

As you’re looking, and that’s even before we got to VR or all the stuff that you were developing and innovating.  As you’re looking 10 years, 20 years down the line, what do you see in terms of the state of the art of the possible that is actually going to potentially reinvent how we fundamentally fight at the end of the day?

Mr. Luckey:  I think that you’re going to see people move away from being the thing that is controlling machines and weapons moment to moment, into being someone who controls things on a much bigger picture scale.

For example, I don’t think that 10, well, probably 10 years from now, but I think 20 or 30 years from now you’re not going to have soldiers that are actually carrying their primary arms.  They might carry some sidearm, but they are mostly going to be relying on autonomous ground vehicles, and autonomous aerial vehicles to be carrying the actual arms. They’re going to be acting as intelligent human target designators on the ground.  That’s just what they’re going to be doing.  And that’s going to be a huge change in the way that we conduct war right now, where a lot of these guys, their job is to carry a metal gun into the right place so that it can shoot the other side.  We’re going to move away from that, and that’s going to be really good for our people that are fighting overseas.

I think Mattis mentioned it in his keynote, that 85 percent of the casualties in the U.S. Armed Forces are in close combat infantry which only makes up I think something like four percent of the force.  So you have this huge disproportionate casualty rate and it’s a very specific group of people.  If we can build technology that is good enough, I think we can bring those casualty rates in line with the rest of the armed forces, and we really should do that.

Mr. Muradian:  How do you operate, though, one of the things our adversaries are working very ardently on is electronic warfare.  And if you look at it, Russia and China both have invested significantly to deny the electromagnetic spectrum to its potential adversaries.  Russia has shown some startling capabilities.

We saw a little bit of it in Trident Juncture in the jamming of GPS signals, but we know from what they did in Ukraine and some of the things that they’ve done to us, that it’s actually much more profound in their capabilities.

How does all of this interconnected universe work where a guy is actually actively working to deny you access to that electromagnetic spectrum?

Mr. Luckey:  So here’s the thing.  Artificial intelligence is going to make us more resistant to jamming, not less resistant. More resistant.  And the reason for that is that right now all these systems that have remote links rely on those links to even operate.  A Predator loses its link and you’re going to have a pretty bad time.

We already have today, in my company we build autonomous helicopter drones that are able to execute on their mission even without a person currently linked to them. So you can say, like our helicopter drones.  You don’t have to remotely pilot them.  You don’t say go forward, go back, point your camera here, point your camera there. You task the on-board AI.  You say go to this place and find all of the white cars.  Go around this perimeter and tell me when something crosses the perimeter.  Tell me when a person is crossing that perimeter.  And even if you jam those links, they can still continue to perform their mission in some capacity.  So if all of our processing is out at the edge, and we have AI integrated with all of these systems, we’re actually able to become less reliant on those remote links.

The other thing that we need to do is to get really good at multi-spectrum communications.  We can’t just do all RF.  We need local communication across acoustics, across the visible spectrum, against infrared.  We need to build systems that are able to communicate data back and forth in multiple ways, because it’s —

Mr. Muradian:  Quantum communication.

Mr. Luckey:  Well, quantum communication would be the ultimate thing.  If we can start entangling bits and throwing the bits on all these things, that would be fantastic.  I think we’re a ways away from that.  Probably at least a few decades from reliable quantum communication.

Mr. Muradian:  Sorry, I got excited —

Mr. Luckey:  No, quantum communication is one of my favorite things.  I just think we’re pretty far away from having really stable entangled bits.  My understanding is they fall apart pretty quickly and we’re not able to keep them entangled.  But when we get that, that’s going to be fantastic.

But in the meanwhile, we can do things like build multi-spectrum links so that even when they try to jam one spectrum you still have a whole of other to rely on.  And it’s very, very hard to jam visible spectrum and acoustic spectrum and the entire EM spectrum all at the same time.  At that point, you’re probably inside of a solid block of rock.

Mr. Muradian:  And the last question.  DIUx. At the time that it was founded there was a lot of criticism that it was focused on well, it’s kind of weird, it’s not properly backed, and then we had 2.0 and had Raj Shah that was dynamically engaging with everybody, and from your standpoint it made all the difference in the world.

Talk to us a little bit about DIUx and what was really, really good about it and what was so influential about it and how it paved the way, from your perspective, for the future for not just yourself but a lot of other people.

Mr. Luckey:  Well DIU, and back when it was DIUx as well, did a great job of proving that people in Silicon Valley could actually get stuff into production, actually do work with the government.  I mentioned to you before this, I don’t think that I would have started this company if it wasn’t for the work of people like Raj Shah doing great work and proving that you actually could get into it.  I would have just assumed it was impenetrable, impossible to get into, impossible to actually make a difference.  And after seeing a lot of the stuff that they’re doing, it’s still early, but I’m a believer that that’s the right way to be approaching this.  And I think that they made a big different already and I think they’re going to become an extremely important organization as time goes on.

Mr. Muradian:  Palmer Luckey, founder of Oculus Rift and now with Anduril, thanks very, very much.  Absolute treat talking to you.

Mr. Luckey:  Thank you.

Mr. Muradian:  I look forward to staying in touch because I think that the more people who are discussing — well, let me ask you one last question.  One last question and then I’ll say goodbye to you again. We’ll have to cut this in.

Raj and I were talking about the rather simplified view of, for example, Google. Right?  I mean a friend of mine was talking about Google and they said okay, so let me see if I get this straight.  China good, Pentagon bad because of what happened with Maven.  But Raj’s point is, actually it’s a lot more sophisticated and that’s not as common as everybody wants to believe, right?  He said that’s a case, but there are a lot of other cases where the whole technology sector is very, very engaged.

What are some of the challenges, though, when you’re looking at very, very transnational companies at the end of the day who have global interests?  What are the right ways to think about that nexus between technology, social, communication, in terms of this ecosystem on which — because the Pentagon is not that innovator anymore.  That was a model that existed because of World War II investment and then it’s petered out, and now the world is going back to the natural order of things where it’s Ford that develops a better vehicle.  It’s not the U.S. Army that develops a better vehicle.  Or the aircraft companies are going to develop the new airplane and bring it to the attention of the Pentagon.

Mr. Luckey:  Yes.

Mr. Muradian:  What’s the right way for the government to think about this space of highly creative people, global interests, folks who want to make a billion dollars because they come up with that great idea, and then go on and make $5 billion with the next good idea?

What’s the right way for leaders who come from a very different background to be thinking about this very dynamic, interesting, potentially conflicted future that we’re going to?

Mr. Luckey:  Leaders need to recognize that there is that conflict.  I think what you touched on, the fact that these are multinational corporations can’t be ignored.  I think if I’m a leader in government I have to think, is this publicly traded, multinational corporation that wants to be in China or is in China, makes potentially hundreds of billions of dollars off of China or wants to in the future, are they really going to have the interests of the U.S. at heart?  Is that going to be their number one target?

I think, unfortunately, in many cases that’s just not so.  So a company has hundreds of billions of reasons to not prioritize the United States military.  You can’t expect that you’re going to be able to rely on them.  And I think that leaders should be taking that into account when they’re looking at these things.  They can’t look at it in a vacuum and say well the headquarters is in America so that obviously must mean these are, that they’re going to be 100 percent behind the U.S. government, behind the U.S. military.

Mr. Muradian:  And what’s the right way — everybody talks about cyber protection. It’s a major issue, whether it’s through persistent penetrations or whether it’s through the theft of some of the most top secret information we have.  And yet everybody talks about it.  But the amount of money that’s actually being invested into protection you could say is grossly inferior compared to the magnitude of the challenge.

Mr. Luckey:  What’s the number that’s going around?  I can’t stand behind this, but I heard that the estimated value is something like $600 billion in stolen IP to China alone.

Mr. Muradian:  Correct.

Mr. Luckey:  And certainly we have not invested $600 billion in cyber protection.

Mr. Muradian:  And a Chinese official once told me, he said look, if it was really that dear to you, you’d protect it more than you do.  Right?  Which I thought was kind of a semi-honest point.  Right?  What is the organismic way to look at the whole challenge if you’re going to build an architecture that’s going to defend it all the way from the lowest level all the way to the highest level in terms of assuring the security because that IP is the coin of the realm?

Mr. Luckey:  You know, I’m going to have to pull my parachute on this one.  I couldn’t possibly say.  I’m not an expert at how you would build an architecture from the top to the bottom across the entire system and protect all of our stuff. This is a combination of technology issues, policy issues, and also willpower.  A lot of these companies that are having their IP stolen, they just don’t really care if their IP is stolen.  I don’t mean that in a flippant way.  I mean literally to them it’s not life and death.  They’re not thinking this is the future of the country or not. It’s like oh, well we got some old IP stolen that we’ve already gotten to market.  Sure, they’re jumping the Chinese ahead five years or ten years in some program. But if you’re only worried about commercial sales, maybe that’s not really all that much of a problem. Especially if you’re not even in those Asian markets where China is.

So there’s kind of this misalignment of incentives at the core of the problem.  The reason the private sector does not invest in cyber security the way that the defense community was is because they’re not thinking of it as a defense problem.  They’re not thinking about this as a fight between great powers.  They’re thinking about it as a fight between corporations so they’re going to invest accordingly.  If it costs more to protect their IP, then they’re going to lose in just the enterprise, commercial sector, they’re probably going to make the rational business decision and decide that they’re just not going to protect their stuff as well.

And that is the real problem that we have to solve.  How can we get these large companies aligned with U.S. interests as a country?  That’s when they will start investing more in protecting their stuff beyond what even makes financial sense just in the market.

Mr. Muradian:  Palmer Luckey, founder of Oculus Rift and Anduril, a fascinating new technology company.  All the best of luck to you.  Thanks very much.  It was a real pleasure.

Exactly.  Live long and prosper, Palmer [Freeman] Luckey.

Mr. Luckey:  Yeah.


Comments are closed.

Your Information will never be shared with any third party.