HEFFNER: I am Alexander Heffner, your host on The Open Mind.

I'm delighted to welcome our guest today, Chris Miller.

He's author of Chip War: The Fight for the World's Most Critical Technology.

Welcome, Chris.

MILLER: Thanks for having me.

HEFFNER: What inspired you to write this book, Chris?

MILLER: Well, I came to realize that our entire world is built on a very fragile foundation of millions and millions of silicon chips, which we never see because they're buried deep in our electronic devices, but we simply can't live without.

HEFFNER: And since you wrote the book, what has changed in this ever-evolving space?

So, within the last year, what are the debates right now occurring within the chip space?

MILLER: Well, the entire world has come to learn just how vulnerable the entire semiconductor supply chain is.

Making chips is the most complex manufactured good that humans produce, and it's just a tiny number of companies in just a couple of countries that produce almost all of the world's most advanced semiconductors.

And the best data point on this is Taiwan, which today produces 90% of the world's most advanced chips.

The type of chip in your smartphone, in many PCs, in data centers, in the telecoms infrastructure, just one company has 90% market share of these critical devices.

HEFFNER: Could they be made in the United States?

What would that look like?

MILLER: Well, it's conceptually possible.

It's just a question of whether the small number of companies that are close to producing cutting-edge chips are willing and able to open up factories in the United States.

Today, the world's leading ship maker when it comes to processor chips is the Taiwanese firm that I mentioned the second firm slightly behind them as a Korean firm, Samsung, Intel in the U.S. has fallen behind a bit in recent years and is racing to catch up.

So, there's no reason why one can't make advanced chips in the U.S., but there's just a couple of companies in this race.

And so whichever company ends up on top will shape which country these chips are made in.

HEFFNER: Why has it not happened in the United States?

MILLER: Well, the U.S. actually invented the chip industry in the late 1950s and early 1960s.

But as chips got more and more complex, the supply chain specialized in the U.S. companies decided to focus on the design of semiconductors.

And so today, companies like Nvidia, Qualcomm, Broadcom, Apple, they design ultra-complex chips, and they make large quantities of money doing so.

But all the manufacturing for these companies takes place offshore, largely in Taiwan.

And so, for the US it's been a good financial decision thus far.

The companies that design chips have often been highly profitable, but it's also created vulnerabilities because the entire U.S. tech sector is reliant on just a small number of factories in East Asia.

HEFFNER: There is a movement in the U.S. to shift away from the foreign production, the outsourcing of this technology.

Politicians talk about it day and night.

So, my question really was why has that not happened yet, since politicians started talking about this in earnest over the last decade?

More frequently over the last, I'd say, four or five years?

MILLER: Well, here's the thing.

Politicians aren't in charge of this semiconductor supply chain.

Companies are, and companies have decided to produce chips where it's the most low cost and a high tech to do so.

And so, for all of the key tech companies right now, whether it's Apple, Facebook, Google, Amazon, they all buy chips above all from Taiwan.

And right now, they're pretty happy to continue this status quo.

The government is trying to change the shape of semiconductor supply chains by incentivizing firms to open up new facilities in the U.S. and via the CHIPS Act, which is administered by the Commerce Department.

Companies can apply for government funding to help to fray the cost of new facilities in the U.S. And that's going to have an impact over time.

But we should be realistic about how large that impact will or won't be.

The reality is that Taiwan is at the center of the chip industry today, and it likely will remain a critical player for a long time to come.

HEFFNER: The argument of those politicians though, is that our autonomy, our digital autonomy, and frankly our liberty as a people is at risk by the foreign production, even by American companies in international outposts.

So, is there truth to this thesis that we are increasingly vulnerable and more insecure as a country every day that these factories blossom overseas, and we don't produce those multifaceted, sophisticated, complex chips ourselves?

MILLER: Well, I think we've got to be specific about words like overseas.

I don't see any risks to producing chips in Japan or producing chips in Germany or producing chips in Korea.

But the last decade has seen, on the one hand, an extraordinary concentration in advanced ship making capabilities in Taiwan.

Just as U.S. military capabilities to defend Taiwan have deteriorated in the face of China's rapid growth of its military power.

And that dynamic is extraordinarily dangerous.

We're more technologically and economically reliant on an island that we're less and less capable of defending, even as Chinese leaders like President Xi Jinping state more and more openly and spend more and more explicitly to build the military that they believe they'll need to take Taiwan.

HEFFNER: And hence we're at a juncture where Taiwan is more and more in focus.

You would say that this is an underrepresented part of the issue of why preserving American independence or at least influence in Taiwan is important.

But what would be your advice, having just studied the history of the chip, regardless of the outcome of any kind of military or civil conflict with China over Taiwan, how can the national security interest of the United States with respect to chip making technologies, how can they be secured in, in the event that there is some conflict?

What would be your recommendation to the American authorities about how to go about preserving whatever American intellectual property might exist in Taiwan?

MILLER: Well, I think the key is to deter a conflict from happening in the first place.

And that's why the trends over the past decade have been so dangerous.

For most of the past half century, the entire world has known if there were a conflict in the Taiwan straits, which side would win.

And today that's become uncertain.

As Chinese power has grown, as China builds more ships, more planes, more missiles and stations, most of them in and around the Taiwan straits, it's become more difficult for the U.S. to credibly deter China from pressuring or attacking Taiwan.

And so, until we can change that military dynamic by building up U.S. capabilities, we're going to have this instability at the center, not just of the U.S.-China relationship, but also hanging over the global economy, because the entire world, the U.S., Europe, Japan, everyone depends on chips made in Taiwan.

HEFFNER: And what percent of the economy have you estimated relies on chips?

What percent of Americans' day-to-day activities relies upon chips?

MILLER: Well, it's hard to find a manufactured good today of almost any value that doesn't have chips inside.

It's not just smartphones or computers, the types of devices that you think have chips, it's also dishwashers and refrigerators and microwaves and coffee makers and automobiles.

And new car will have, on average, a thousand semiconductors inside of it.

Today, we can't make almost anything that doesn't involve not just one, but often dozens or hundreds of chips inside.

So, the entire world economy is fundamentally today dependent on semiconductors, which is why the dilemma of the concentration of semiconductor production today is so important, not just for the tech sector, but for the whole economy.

HEFFNER: How do you compare this to natural resources and the kind of historic debates and conflicts over water and over gas and oil.

How does this confrontation compare to those prior confrontations in history?

MILLER: Well, there's almost no resource that is as concentrated as the chip industry is concentrated.

We think of Saudi Arabia as a big player in the oil industry, but the Saudi's produce 10 or 15% of the world's total oil production.

Taiwan produces 90% of the world's most advanced processor chips.

So, it's even far more important for the chip market than Saudi Arabia is for the oil market.

And we've seen over the past half century the way that countries like Saudi Arabia, Iran, other big oil producers, can swing international politics and cause huge disruptions to the world economy by changing the amount of oil they supply to world markets.

But the same is true even more so when it comes to the concentration in chipmaking, which is why as we think about the political implications of this concentration manufacturing we need to understand that this isn't simply an economic question, it's also a geopolitical question.

HEFFNER: It's geopolitical, and as you teach the next generation of students about semiconductors and chips and their relevance in a potential World War III, worst case scenario, what are you advising is the best way to negotiate this through diplomacy with the players that we know are in many cases immutable with respect to authoritarian tendencies and just the way the world is right now?

Unlike the lofty aspiration of my friend Francis Fukuyama, there was no heartbeat of democracy around the world, kumbaya, let's all get along.

And in the current climate of powers like Russia and China being challenged either, militarily or economically, but especially economically, with the existing players, how do you engage in a constructive American strategy that does not involve combat?

MILLER: Well, I think right now there's a race in the, the tech sector race between a Chinese led system and a U.S. led system that involves partners like Taiwan, like Japan, like Korea, to harness the most advanced technological capabilities and deploy them to all sorts of different use cases.

But today, the most advanced chips are being used to train Artificial Intelligence systems.

And so, it's really the AI race.

That is why semiconductors are so important technologically today.

And again, in the training of AI system, there's also extraordinary concentration.

Most of the chips used to train AI systems are designed by just a couple of companies, almost all U.S. based, and most of them are manufactured again in Taiwan.

And so that's why chips are at the center of this geopolitical race with economic ramifications, but also with military ramifications as governments try to apply AI to military systems.

HEFFNER: What I hear you saying, and it makes sense in the context of a plutocratic society and a corporate cannibalistic society in many respects, is that the way that corporations function going forward is going to determine the fate of the civilized world or civil society, not necessarily the politicians.

That that's what I hear you saying, that corporations are going to have more power over how this ends than maybe the voters or the people the voters elect or the totalitarians who rule over the voters.

MILLER: Well, I would look at it this way.

Today, 98% of chips that are produced go to commercial applications rather than government applications.

The government just buys 2% of chips.

And so, the incentives of the commercial marketplace will be the dominant factor in shaping chip supply chains, who produces the most events, semiconductors, where that takes place.

And so, companies like Apple, like Nvidia, like Qualcomm, like Intel, they will shape the future of the chip industry.

And so, the interests of their consumers, do their consumers care where their chips are made what security concerns do their customers have?

These are the factors that are going to play the biggest role in shaping where semiconductors are manufactured.

HEFFNER: So if that's the case and how these companies are programming AI into their systems then, then that's the ball game.

That's going to determine the fate of peace or violence or whether countries that are authoritarian now transform into more democratic systems.

Let me ask you this, having studied this up close is the way companies that are operate or based in democratic societies, is the way that they are programming AI significantly different from the way that AI is being programmed by companies that are operating in non-democratic countries?

MILLER: Well, I think we're in the early stages of AI, so it's still hard to draw definitive conclusions, and every company is drawing up their own set of guidelines.

Different countries are bringing different regulatory systems to bear, but I think we should assume, you know, the history of technology shows that the ways technologies are deployed are inevitably shaped by the societies and the political systems that they're operating in.

And I don't think AI will be any different.

Different societies will deploy AI in different ways shaped by their political and social features.

HEFFNER: My friend Joy Buolamwini has come on The Open Mind a few times to talk about the threat internally, how AI has been devised in our own country to injure the quality of life of a segment of society primarily Black and Brown people, but underserved people, economically disenfranchised people.

That's why I'm asking that question because we have seen that exposed in the United States how AI has gotten off to a rocky start in representing the ideal of life, liberty and happiness for everybody.

But I would think that as we're involved, albeit not with our U.S. troops on the ground, in Russia's invasion of Ukraine, and what continues to be perceived as possibly the beginning of a Third World War or major global conflict, I would think that we would have some intelligence now about how AI systems are operating if they are in Russia, but also in certainly in China.

Is there any light you can shed on that?

MILLER: I think AI is just a tool, right?

It's going to be used by, in different ways, by different actors.

Companies will use it in certain ways.

Governments will use it in other ways.

We know that in particular in China, there's lots of AI activity in the corporate space.

We know that Chinese government is trying to acquire AI capabilities for its military.

We know it's already used AI capabilities like facial recognition in a widespread sense for surveillance.

I don't really think there's good or bad AI, I'm good or bad uses of, of ai, and we should expect bad actors to use it badly.

HEFFNER: But it's not that simplistic, right?

I mean, in the sense that there are ways that AI can be programmed to try to wipe out human relevance, right?

And, in that particular vein, when these companies are thinking about the disposability of the human being, not just from the economic perspective but maybe from some “Total Recall ” perspective, right?

The idea of AI taking over, is there a fear in reporting the book and your ongoing evaluation of this issue that AI could be poised to usurp control over human decision-making you know, on important things that that could kind of make-or-break war and peace.

Are these companies at a point where they're nervous about what they're empowering the AI to do?

MILLER: I think there are people who are nervous about these capabilities, but I guess I have to say that in, in what I've seen in the cases where AI is set up in a responsible way I don't see a lot of reason to be concerned about the doomsday scenarios, and did I struggle to really understand what the doomsday scenarios actually are?

The past couple of years we've seen some pretty cool toys like Chat GPT that can finish sentences for us and do a better job at Google search than a better job at search than Google can.

But I think it's a big jump from there.

HEFFNER: I'm going to stick with my Google algorithm of discriminating against all non .edu websites.

I wonder if there's going to be Chat GPT intellectual academic edition that can live up to the JSTORs and the more discriminating tactics of digital navigators.

But what I'm getting at it, you alluded to it broadly, the idea that AI decision-making could become more precise, and when it involves a question of human values or moral values that there are those who would put that decision authority in the hand of AI over the human being, and they may make a cogent persuasive argument about doing that.

That's really what I'm getting at, right?

MILLER: That's right.

And there are certainly cases where we, we should and should not feel comfortable putting decisions in the hands of machines.

And so it's our job as ethical actors to decide which cases we're comfortable, which cases we're not comfortable.

But that's true for all types of technology.

The specific context are different with AI, but the dilemmas of where do we trust machines, where do we not trust them, that's an old dilemma in moral and ethical philosophy.

HEFFNER: Right.

Is there anything after chips?

Because we think of like Google still as the new and last frontier, right?

We, we kind of have felt like Twitter now X and Facebook primarily focused on Instagram now and WhatsApp, like they are the first and last frontier in the social space.

And what comes after chips?

MILLER: Well, I think for a long time nothing's going to come after chips.

We're going to be living in a silicon dominated world for many, many years to come.

People talk about quantum computing, which at some point will probably materialize.

But until then, and probably even well into the quantum era, we're going to be reliant on thousands and thousands of silicon chips that structure our economy, structure, society structure our daily lives.

HEFFNER: I asked you to make a recommendation to the American government, but what is your sense of how well we are guarding the technology?

Like to the extent that we have databases that represent what we're doing abroad and at home, do you feel like the cybersecurity networks of this country are adequately protecting the knowledge about what chips we do possess and what their capabilities are?

MILLER: The last couple of years, the government has made a lot of efforts to increase its expertise in this sphere.

There's still a lot of work to be done, but the level of knowledge and information is much higher than it would've been just a handful of years ago.

HEFFNER: Okay.

Take care.

MILLER: Alright, thank you for having me.

HEFFNER: Please visit The Open Mind website at thirteen.org/open Mind to view this program online or to access over 1500 other interviews.

And do check us out on Twitter and Facebook at Open Mind TV for updates on future programming.

Continuing production of The Open Mind has been made possible by grants from Ann Ulnick, Joan Ganz Cooney, Lawrence B Benenson, the Angelson Family Foundation, Alfred P. Sloan Foundation, the John S. and James L. Knight Foundation, William and Flora Hewlett Foundation.

JoAnne and Kenneth Wellner Foundation.

And from the corporate community, Mutual of America.