In America’s Next War, Machines Will Do the Thinking

In America’s Next War, Machines Will Do the Thinking

Think of all the things we depend on daily life that were made possible by the U.S. military: the internet, jet travel, GPS and, perhaps most important, duct tape. That such things grew out of a desire to find new and better ways to kill lots and lots of people is easy to put out of mind, just as we tend to forget where that juicy ribeye on our plate originally came from. 

If consumers want truly smart machines (smarter than Amazon’s new Astro, anyway) and the “internet of things,” Google and Apple and Elon Musk will all play their roles, but much of the research and development will come from the Pentagon — and, no doubt, the People’s Liberation Army. Congress, holder of the strings to the Defense Department’s $740 billion purse, is going to have to make it happen. Where should the smart money go?

To answer that question, I talked with somebody who knows every nook and cranny of the defense budget, Christian Brose, author of “The Kill Chain: Defending America in the Future of High-Tech Warfare.” Brose spent a decade at the Senate Armed Services Committee, including four years as staff director and top adviser to the chairman, Senator John McCain. He is now the chief strategy officer at Anduril, a defense technology start-up. Here is a lightly edited transcript of our discussion:

Tobin Harshaw: Let’s start with recent news: The U.S.-U.K.-Australia sub deal, knows as Aukus, has the French very upset. What do the U.S. and the Aussies gain from it?

Christian Brose: The prime mover was the desire of the Australians to move away from their existing contract for French submarines, irrespective of what they moved to next. Once that decision was made, then the question becomes, “What is the new capability going to be?”

The trilateral partnership makes a ton of sense. The Australians need nuclear-powered boats to get the ranges and endurance necessary. The other strategic rationale is collaboration in advanced areas such as artificial intelligence and autonomous systems. That is essential for the three allies together to start getting to the kind of scale they need to compete in the Indo-Pacific, which is big, and against China, which is big.

TH: In terms of working with allies in the region, how does this deal fit in with the so-called Quad of Australia, Japan, India and the U.S, and with America’s bilateral relationships with Japan, Taiwan and others?

CB: I’ll be interested to see where the administration takes it. I think before it starts a broad expansion, it needs to plumb the depths of what’s really possible in a common defense-industrial and technological base among the three allies: what we could do together; the ways in which we can collaborate and co-develop new capabilities and systems that are interoperable. The U.S., U.K. and Australia have a unique bond that’s going to be hard to replicate with other countries. So it’s more a set of concentric circles of allies that radiate out from that core group.

TH: Is this then the right model for the new Cold War? Some talk of an “Asian NATO” has been thrown around.

CB: I disagree with the framing of competition with China as a “new Cold War.” So many of the historical analogies don’t hold up. You don’t have geographically divided blocs of countries. As for the allies, there are a lot of reasons why an Asian version of the North Atlantic Treaty Organization isn’t right or achievable, at least not now. New ties in the Indo-Pacific are probably going to grow more organically from the groupings of different sorts of U.S. multilateral arrangements, some of which will be very focused on competing with China, some of which will have other, broader purposes.

TH: Is it understandable perhaps that China feels encircled by deals like Aukus — Foreign Minister Wang Yi on Tuesday said the sub deal brings “hidden dangers into regional peace.” Does it lead to a danger of China lashing out recklessly?

CB: I can’t claim to know what China feels. My contention is that China is pursuing a strategy of displacement, as the analyst Rush Doshi calls it. It is actively trying to displace the U.S. and the current international order — to replace it with some kind of a new order. It’s unclear what that’s actually going to be, how it would function and what its values would be. But I think a lot of countries, especially in Asia, are deeply concerned about what that would mean for them.

TH: Your old shop, the Senate Armed Services Committee, has approved the fiscal 2022 Pentagon budget at around $740 billion. Is that the right amount? Is it being spent on the right things to, as you put it in “The Kill Chain,” buy deterrence?

CB: It’s not about how much we’re spending. The primary question is: Are we spending enough on the right things? The answer is, decidedly not. The defense establishment has been systemically disrupted over the past generation. We’ve been disrupted by peer competitors, primarily China, that have been implementing strategies to undermine the ways in which the we project military power and compete internationally. And we’ve simultaneously been disrupted by technology, primarily these advanced and emerging technologies like artificial intelligence and autonomous systems, which are becoming pervasive in the commercial world yet not nearly fielded to the degree that they should be in the defense world.

TH: How do we deal with these simultaneous challenges?

CB: When you’re in a period of disruption, the worst thing you can do is cling to the status quo. That is playing a losing game. We could spend significantly more money on defense, but if we continue to spend it in the ways that we’re spending it, we still lose. If we made better choices about the ways in which technology and capabilities work or should work together, I think we could build a force more capable of generating deterrence than the one that we have now, for less money.

A lot of the fights around spending are going to come at the expense of time and space of people who need to be thinking about how to reshape our military. I lived this dream for the better part of 10 years in the Senate. We spent thousands of hours negotiating and strategizing over the top line, that $740 billion. And that was time that we never got back to think about the far more important issue — the kind of military that we’re going to need in the future and how to deliver it.

TH: A couple of years ago, I did one of these Q&A’s with Bob Work, the former deputy defense secretary, who was the driving force behind the modernization initiative known as the “third offset.” Did he have any lasting effect, or did third offset end up being nothing but a slogan?

CB: The third offset was the right idea. It recognized the strategic reality. But the implementation has been spottier. For the past five or six years, the Department of Defense has been on an innovation quest. Yet so much of it is turning into “innovation theater.”

We have plenty of innovative people, particularly down at the lower levels of our industry, the military and elsewhere. The problem is that so many of those new and disruptive ideas and concepts and capabilities never get to scale. They are small pilot projects. They’re prototypes. They’re demonstrations. But the best of them do not actually get separated from the herd and very quickly taken to scale so that they can make an impact. Most fail to transition and perish in what’s known as the “valley of death.”

The government has created a plethora of innovation units and organizations. We have tried to change the culture of defense with more hoodies and relaxed grooming standards. But at the end of the day, this is about being able to differentiate the most promising capabilities, the most disruptive capabilities, and getting those to scale fast. And the people who are best positioned to provide those solutions — engineers, companies, founders, investors — are very quickly losing hope that this time it’s for real.

Many have decided to try their hand at working with the Department of Defense, which has been made easier, but there are still precious few success stories. If that continues, a lot of this pool of new entrants is just going to go take its talent elsewhere, which is how we got into this mess in the first place. Many of those folks went off to work in the commercial market, not because they were unpatriotic, not because they didn’t care, but because they just could not see a path to be successful working in national defense. And unfortunately, 25 years of empirical evidence suggests that they’re right.

TH: Where do you think that the Pentagon most needs to engage these innovative companies and people? Is it AI? Unmanned systems? Communications networks?

CB: All of it, but primarily autonomy. Autonomy — enabled by artificial intelligence, edge computing and other technologies — allows you to operate at scales and speeds that you simply cannot do under the traditional model of big, expensive, heavily manned systems that, no matter how much money you throw at them, will only be able to do a limited number of things. The problem is that China has been developing very precise capabilities to disrupt, disable, degrade and destroy those limited numbers of big things. This is our strategic problem. This is why our future force needs larger numbers of cheaper, more autonomous systems.

We’re not talking about photon torpedoes and intergalactic space travel. We’re talking about systems that are already in existence; that are imminently fieldable if we move with the right sense of urgency.

TH: In your book, you wrote a bit about the ethics of this and taking the human out of the “kill chain.” Is that something that’s going to happen?

CB: Yes, in time human beings will need to rely more on intelligent machines to help them understand, decide and act in warfare. And no one should be under any illusion that the highly manual ways we do this now are optimal. Look at the recent drone strike in Afghanistan that killed 10 civilians. There were a lot of humans in a lot of loops throughout the entirety of that strike, and we still made a mistake because humans are imperfect. Let’s not lionize the situation that we have today. The real question to me is what can machines do to enable human beings to work better, to work more effectively, to make better decisions? If AI-enabled systems and autonomous systems can improve those processes, if they can help people minimize mistakes and reduce the risk to civilians and our own forces, that is a morally good thing. Still, accountability is going to have to lie with a human being.

TH: As we saw with Project Maven, where Google dropped out of a Defense Department initiative because of employee protests, sometimes the tech world is less than eager to work with the Pentagon. So if there needs to be greater public-private cooperation on AI systems and autonomous systems, is that Silicon Valley hesitancy going to become a bigger problem?

CB: Large technology companies are not going to ride to the rescue because they’re not incentivized to do real national security work. It is contrary to their business model and even their identity. Some of them may be convinced to do limited things, like sell cloud services, but they’re going to be the same cloud services more or less that are available in the commercial market.

The real question is whether the Department of Defense will create the incentives for the emergence of new defense companies. We can make defense a lot more like the commercial market. But you’re only going to get so far with companies that are seeking to sell the same capabilities that they’ve sold in the commercial market into the defense market. Ultimately, you’re going to need new and different companies to step in and try to do this new and disruptive defense work. In my experience, there are many engineers, founders, companies and investors that would like to do that work. I work at one of those companies, Anduril Industries, but there are others, and ultimately we are going to need a lot more.

And I believe it is possible. Many Americans in technology are open to working in defense because they’re patriots and they believe it’s the right thing to do. Many think that it could be a path to becoming successful, wealthy, famous. And many are just drawn to hard problems, and that’s why they became engineers in the first place. The reason so little of that has happened over the past 25 years is because the government has not created the incentives for these people to succeed in doing this work. The Department of Defense needs to get this right at a structural level, or all of this rhetoric about change and innovation is just going to be hot air.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Tobin Harshaw is an editor and writer on national security and military affairs for Bloomberg Opinion. He was an editor with the op-ed page of the New York Times and the paper’s letters editor.

©2021 Bloomberg L.P.