In the opening years of the Cold War, the world was one misunderstanding away from a nuclear war. But eventually the United States and the Soviet Union entered arms-control talks and developed a set of shared expectations. The stabilization of the decades-long standoff became possible in part because of the comparative simplicity of a bipolar conflict between two superpowers.
In the years since the fall of the Soviet Union in 1991, an era that Yale SOM strategy expert Paul Bracken calls the second nuclear age, the world has become multipolar, with a growing number of nuclear powers, including Pakistan, North Korea, India, and Israel, joining Russia, China, and the United States and its Cold War allies the UK and France. The increasing number of power centers has added significantly to strategic complexity and the danger of conflict, Bracken says.
Bracken also warns against focusing exclusively on nuclear weapons. Today, strategic advantage often flows from information: data from drones, satellite images, cell phones, license plate cameras, all ingested and appraised by artificial intelligence. The demand for such information, and the power that comes with it, is driving a technology arms race.
Q: What is the second nuclear age?
I define the era after the Cold War as the second nuclear age. Rather than two dominant superpowers, we have many powers with big GDPs, advanced technology, and in many cases nuclear weapons. The different world views and different problems facing these powers lead to a different pattern of relationships.
Q: What was the role of technology in shaping each age?
The Cold War and the second nuclear age are both dynamic. Within each period the problems change over time. When the Cold War began in the late 1940s and into the 1950s, people simply did not understand the new technologies. I don’t mean just nuclear weapons. Let’s recall that jet aircraft, faster missiles, nuclear submarines, radar, and eventually satellites were all so new that while you knew who the enemy was you didn’t know how they were going to use the technologies.
“How do you know that video from a drone, data from an automatic license plate reader, a satellite picture, and audio from a hacked cellphone could all contribute to understanding the same issue? This is where artificial intelligence comes in.”
Most Cold War experts conclude that the chance of an accidental launch or a hair-trigger firing of a weapon was much higher in that first decade. In the 1960s and 1970s, as the two sides got to know each other, they built up a set of expectations which wouldn’t guarantee security but would tell you if there was going to be some sudden eruption. This added to stability.
Information technology has been critical to the second nuclear age. Today, artificial intelligence, cloud computing, data analytics, and cyber warfare are all transformative technologies. There’s such tremendous collection of information going on—think of cyberattacks, hacked phones, automatic license plate readers, and on and on—there’s simply no comparison between the information systems of the Cold War and today. We were really blind back then. Now, we’re not blind—at least in peace time. We don’t know how these systems would function in war time.
Q: Your book The Command and Control of Nuclear Forces helped focus attention on the importance of information flows as a part of nuclear strategy. What was the approach when you published the book in 1983?
People focused on the number of nuclear weapons, but they wouldn’t address how they were managed, what information would be available to decision makers in the White House and in the Kremlin during a severe crisis or an actual war. A lot of people, including myself, concluded that the system would, at an informational level, very likely fall apart. Leaders would only have access to very poor or even entirely inaccurate information.
Understanding that led to a shift of strategy. The Reagan administration conducted some war games inside the Pentagon, which I was involved with, that convinced the secretary of defense to redirect the technological thrust of the United States to conventional forces and away from nuclear systems, which required very sophisticated information processing to conduct a nuclear war.
Q: Ostensibly, the Cold War and the second nuclear age are both defined by nuclear weapons. Has that technology changed?
There’s been some innovation of the nuclear technologies themselves, but it’s much less important than the innovations in the surrounding information systems used for targeting and to move weapons around to avoid enemy detection. Is a country preparing weapons for launch? Is North Korea moving their mobile missiles from one province to another? The information advantages are more important than the weapons themselves.
But it’s not just information; it’s being able to manage the flow of real-time, diverse, voluminous information. How do you know that video from a drone, data from an automatic license plate reader, a satellite picture, and audio from a hacked cellphone could all contribute to understanding the same issue? This is where artificial intelligence comes in. No human mind, no military staff could manage this amount of information. That was one of the big drivers behind China’s decision to become the world leader in AI.
When you have these tools, you end up with very subtle and creative possibilities. You can track the license plate of the motorcycle of a member of a particular unit in the Pakistani army’s nuclear force. You can hack into the security cameras in Islamabad, which there’s good reason to think Huawei has done. When he gets an alert, which is intercepted, saying, “Report to your station in the next two hours,” you have valuable intelligence. That level of information was unthinkable even a few years ago.
An information advantage changes the military balance. I would say China is in the lead on information. The U.S. is now responding vigorously in various ways. It’s a dynamic state of affairs.
Q: And these new technologies are being introduced into a complex, multipolar world.
It’s worth noting: the bipolar world of the Cold War was rare historically. Through most of recorded history, you had multiple decision-making centers with significant military and economic power. The world has normally been multipolar, particularly Europe and Asia. Even during the Cold War, the idea that it was a bipolar world was, to some extent, a useful fiction because let’s not forget that the French, the British, and after 1964, the Chinese all had nuclear weapons, too.
The point is a multipolar world is more complex. Let me just define complexity because it’s one of the most overused, undefined terms out there. All I mean by complex is there are many moving parts, many decision-making centers. The increase in the number of powers is significant. One of the big divisions in game theory is between simple game theory, which has two players, and n-person game theory where n is any number of players. We have nine states with nuclear weapons. We have 18 states that host nuclear weapons or other critical infrastructure for nuclear systems.
The single biggest lesson that game theory teaches is that coalitions form as soon as you have more than two players. The U.S. can align with Germany, Great Britain, and India. Russia can align with China, possibly with North Korea and Pakistan. You have coalition dynamics. How do we keep our coalition together? How can we drive a wedge into our enemy’s coalition so that it’s not as stable?
Nixon’s efforts to open China were part of the U.S. effort to drive a wedge in the communist bloc. The role that nuclear weapons played in this is not that well known. When Richard Nixon went to Beijing in 1972, his assistant Henry Kissinger had in his briefcase the locations, longitude and latitude, of every Soviet nuclear weapon. We gave this information to the Chinese in the Great Hall of the People in Beijing as a sign of good faith. It gave the Chinese information they could never possibly have collected on their own.
That information transfer was instrumental in changing the coalition. It probably made China a less reliable ally of Moscow. It’s likely that Russia increasingly saw China as in alliance with the United States. The whole point was to keep the Soviets from planning a war on Europe without worrying what might happen in Asia.
Q: How does this research show up in your teaching at Yale SOM?
There are easily 15 technologies that are critically important to national defense strategy. That number is so large that the Pentagon doesn’t know which ones to focus on. Moreover, they think of these technologies in isolation from each other. One course that I teach at SOM focuses on the need to think about technology packages—two or three technologies integrated with each other.
“The U.S. and China are not going to go to war. We both have too much to lose. But it’s all too easy to imagine some general seeking power setting off a change of government in North Korea or a war between India and Pakistan.”
Where there’s a breakdown, as I see it, is in the area of leadership. There’s a tremendous amount of innovation but it’s bottom up, not top down. These technologies are developed and supplied by contractors and engineers with specialties in the various technologies. What we don’t do a good job on in the United States is to establish a direction—that is, provide leadership from the top. There’s value in offering guidance to build technology packages to do certain things and not other things. That’s actually true in the corporate world as well as the DOD. If the students at SOM are any indication, there’s a real desire to learn technology management and leadership.
Q: Where do you see the multipolar world going?
We can think about plausible futures looking out 10 years. One is what happened in the first part of the Cold War, which is to say, everybody wants these technologies, so there’s a kind of mad catch-up attitude that leads to a technology arms race.
If everybody goes down that road, we’re going to have a very unstable system. We went through that during the Cold War. The paranoia about communism in the 1950s meant any discussion of cooperation with Moscow on arms control was dismissed out of hand. By 1965, there was a complete reversal. We were all about arms control and building up relations with the Soviets so that we didn’t misunderstand each other and get into a disaster.
The cycle is likely to repeat itself. But there needs to be a focus on second-tier nuclear states like Pakistan and North Korea, perhaps Iran, perhaps Israel. The U.S. and China are not going to go to war. We both have too much to lose. But I think Pakistan and North Korea are fundamentally unstable. It’s all too easy to imagine some general seeking power setting off a change of government in North Korea or a war between India and Pakistan
Right now, China is focused on us, but Pakistan is on their border. North Korea is on their border. It doesn’t seem to have dawned on the Chinese yet that these countries are more likely than the U.S. to be a source of a catastrophe that they don’t want to get sucked into.
Q: Should we be thinking about arms control negotiation with the Chinese?
I came out of the think tank world, where the biggest single difficulty is to break out of the current mindset. Today, all anyone can think about is coronavirus, no matter what you are studying. In 1978, all anyone could think about was the energy crisis. In 1956, at the height of the Cold War, someone who said “I want to see if we could stabilize the arms race through an arms-control agreement with the Soviets” would get laughed out of the room.
When I suggest U.S.-China arms control today, the response is similar. I guarantee you it will be something we are talking about in six or seven years which is precisely why now is a good time to think about it, because there’s no pressure. Nobody’s going to take it seriously. You may be ridiculed but it’s important to do the foundational work before it’s needed.
In my course on problem framing, I continually say, you can’t focus on the most likely outcome; you’ve got to focus on what could do you the most damage and, alternatively, what could lead to the biggest opportunity.
Don’t tell me what is going to happen; look at the full band of possibilities. Give me scenarios A, B, and C. And if C means we’re in real trouble, tell me how we come up with a plan beforehand to make sure that it’s not too disastrous.
Too often, decision makers hear “unlikely” and immediately reduce a small probability to zero, then pay it no attention whatsoever. We’re living through a case study of that. Preparing for a pandemic is a lot more effective and less expensive than dealing with one after it arrives. On the positive side, I think COVID-19 will lead people to appreciate the importance of not focusing entirely on the most likely case. Particularly in leadership roles, they’re going to look for a band of possibilities.
Interview conducted and edited by Ted O’Callahan.