24 – Chip War: Mastering Tech, Geopolitics & the China Challenge

WEBVTT
00:00:00.000 --> 00:00:05.560
Welcome to Berlin Side Out, the podcast that takes an expert look at how Germany sees the
00:00:05.560 --> 00:00:07.840
world and the world sees Germany.
00:00:07.840 --> 00:00:09.600
With me, Benjamin Tallis.
00:00:09.600 --> 00:00:11.840
And me, Aaron Gash-Bernett.
00:00:11.840 --> 00:00:20.180
Hello, and welcome back to Berlin Side Out, the Foreign Affairs podcast in association
00:00:20.180 --> 00:00:25.880
with the German Council on Foreign Relations that takes an expert look at how Germany sees
00:00:25.880 --> 00:00:27.760
the world and the world sees Germany.
00:00:27.760 --> 00:00:33.000
I'm Aaron Gash-Bernett, a journalist and security analyst specializing in German politics and
00:00:33.000 --> 00:00:34.240
foreign policy.
00:00:34.240 --> 00:00:38.680
And I'm here with my friend and co-host, Benjamin Tallis, senior research fellow here at the
00:00:38.680 --> 00:00:42.400
Council and head of its action group, Seitanbende.
00:00:42.400 --> 00:00:46.800
Now Ben, regular Berlin Side Out listeners will know how important we think it is to
00:00:46.800 --> 00:00:51.720
ensure that we're all first at least aware of the systemic competition that we're currently
00:00:51.720 --> 00:00:55.080
in, in between liberal democracies on one side.
00:00:55.080 --> 00:01:00.140
That includes Germany and authoritarian regimes like Russia and China on the other.
00:01:00.140 --> 00:01:04.280
And that's because from that awareness, we can act strategically and proactively rather
00:01:04.280 --> 00:01:09.640
than reacting to an authoritarian set agenda from the back foot, which we unfortunately
00:01:09.640 --> 00:01:12.080
too often do.
00:01:12.080 --> 00:01:17.600
Now that systemic struggle has military, economic and soft power fronts and having a plan for
00:01:17.600 --> 00:01:22.600
how to use all those resources and power that democratic states have is part of the grand
00:01:22.600 --> 00:01:23.600
strategy.
00:01:23.600 --> 00:01:28.440
We've spoken about regularly on this podcast, including in detail a few episodes ago with
00:01:28.440 --> 00:01:31.840
Maximilian Tehalla, Eileen Matler and Jakob Brass.
00:01:31.840 --> 00:01:37.000
But there is a technological dimension to these strategic questions too, isn't there?
00:01:37.000 --> 00:01:38.160
Yes, there certainly is.
00:01:38.160 --> 00:01:42.120
And living in Germany, we're acutely aware of the need to master technological change
00:01:42.120 --> 00:01:46.120
if not quite the reality of doing so at the moment.
00:01:46.120 --> 00:01:51.320
And our guest today, Chris Miller, has done anyone who works in geopolitics an enormous
00:01:51.320 --> 00:01:58.480
favour by making clear in the book, "Chip War," the interlinkages between technology, geopolitics,
00:01:58.480 --> 00:02:04.320
geoeconomics, between the different parts of our societies and how absolutely crucial
00:02:04.320 --> 00:02:08.480
the access to the technology, particularly semiconductors, that power our lives and
00:02:08.480 --> 00:02:10.040
our economies is.
00:02:10.040 --> 00:02:15.000
But moreover, to highlight the technological aspects of the race for power, the race for
00:02:15.000 --> 00:02:19.040
more advanced technology, for the technology that keeps us safe and gives us a military
00:02:19.040 --> 00:02:20.040
edge.
00:02:20.040 --> 00:02:24.120
And it's a long history that Chris traces in wonderful detail through the Cold War
00:02:24.120 --> 00:02:28.680
and into the post-Cold War period, right up to the present day.
00:02:28.680 --> 00:02:32.720
And the systemic competition with China and also with Russia that we are engaging in.
00:02:32.720 --> 00:02:38.600
And it's those intertwined histories of economics, politics, science, technology and military
00:02:38.600 --> 00:02:43.640
affairs that really fascinated me about this book, which I can't recommend highly enough.
00:02:43.640 --> 00:02:48.200
Rarely have I learned so much from a book I've enjoyed so much, I have to say.
00:02:48.200 --> 00:02:51.960
And so to all our readers out there who have yet to have the pleasure, you have a treat
00:02:51.960 --> 00:02:52.960
ahead of you.
00:02:52.960 --> 00:02:54.440
But it's not just me who thinks that.
00:02:54.440 --> 00:02:58.320
The book "Chip War" has been named "Business Book of the Year" by the Financial Times.
00:02:58.320 --> 00:03:01.240
The Economist called it one of their book of the years.
00:03:01.240 --> 00:03:03.840
As did Foreign Affairs, it's a New York Times bestseller.
00:03:03.840 --> 00:03:04.840
I could go on.
00:03:04.840 --> 00:03:08.880
The list of awards and accolades for Chris's work is long and distinguished.
00:03:08.880 --> 00:03:13.080
Chris Miller is professor, an economic historian at the Fletcher School, the Global Affairs
00:03:13.080 --> 00:03:16.640
Graduate School at Tufts University, as well as author of "Chip War," as we mentioned,
00:03:16.640 --> 00:03:17.640
and several other books.
00:03:17.640 --> 00:03:18.640
Do look him up.
00:03:18.640 --> 00:03:21.920
Chris, it's a pleasure to see you again and to have you on Berlin's side out.
00:03:21.920 --> 00:03:24.200
Thank you very much for the invitation to join you.
00:03:24.200 --> 00:03:25.880
Thank you very much for joining us, Chris.
00:03:25.880 --> 00:03:30.160
Now, many of our listeners might be familiar with China's historical claim, which it's
00:03:30.160 --> 00:03:34.200
had since 1949, that Taiwan is a province of China.
00:03:34.200 --> 00:03:39.280
It has no right to be its own state, and that it doesn't rule out force to take Taiwan.
00:03:39.280 --> 00:03:43.920
Now, our listeners are also no strangers to the dangers of authoritarian states, threatening
00:03:43.920 --> 00:03:48.480
democratic states motivated by imperial history.
00:03:48.480 --> 00:03:52.920
But your book brings up another particularly noteworthy reason why China might want to
00:03:52.920 --> 00:03:54.680
invade Taiwan.
00:03:54.680 --> 00:03:58.280
As you argue, chips are the new oil.
00:03:58.280 --> 00:04:05.640
This is a big central point that you make in the book, and Taiwanese chip manufacturers
00:04:05.640 --> 00:04:09.560
lead the world right now in terms of manufacturing know-how.
00:04:09.560 --> 00:04:14.760
So let's start by asking just how dominant is that Taiwanese know-how in terms of that
00:04:14.760 --> 00:04:20.320
manufacturing, and is that a reason why we might be under appreciating the risk of China
00:04:20.320 --> 00:04:22.480
trying to invade Taiwan potentially?
00:04:22.480 --> 00:04:27.560
Well, it's definitely the case that Taiwan is the most important country when it comes
00:04:27.560 --> 00:04:30.400
to manufacturing chips.
00:04:30.400 --> 00:04:36.000
There's a Taiwanese company called TSMC, the Taiwan Semiconductor Manufacturing Company,
00:04:36.000 --> 00:04:40.000
which is the world's largest chip maker, and also the world's most advanced chip maker
00:04:40.000 --> 00:04:43.040
in terms of its ability to produce processor chips.
00:04:43.040 --> 00:04:45.280
And advanced processor chips are basically everything.
00:04:45.280 --> 00:04:49.960
They're in your phone, computer, they're in the telecoms infrastructure, they're in the
00:04:49.960 --> 00:04:55.200
data centers that are increasingly important for training and deploying artificial intelligence
00:04:55.200 --> 00:04:56.200
systems.
00:04:56.200 --> 00:05:01.280
And when it comes to the most advanced chips, 90% of them are made by TSMC in Taiwan.
00:05:01.280 --> 00:05:07.960
And when it comes to AI processors, almost 100% of them are made by TSMC in Taiwan.
00:05:07.960 --> 00:05:11.760
And that's true for AI processors that are used in Asia, used in Europe, used in the
00:05:11.760 --> 00:05:13.840
US, even used in China.
00:05:13.840 --> 00:05:18.880
And so that means that Taiwan is really at the center of the world's technology supply
00:05:18.880 --> 00:05:19.880
chains.
00:05:19.880 --> 00:05:24.320
Most people don't realize it, but for almost every aspect of their daily lives, they rely
00:05:24.320 --> 00:05:27.720
on TSMC chips whenever they touch a piece of technology.
00:05:27.720 --> 00:05:33.120
And so it really is hard to understate Taiwan's economic and technological importance to the
00:05:33.120 --> 00:05:34.120
entire world.
00:05:34.120 --> 00:05:35.120
Right.
00:05:35.120 --> 00:05:39.080
What you bring out very clearly in the book is the salience of that particular node in
00:05:39.080 --> 00:05:43.160
what I think you rightly described as the incredibly intricate, incredibly complex and
00:05:43.160 --> 00:05:48.280
efficient, yet incredibly vulnerable global supply chain that we've managed to build for
00:05:48.280 --> 00:05:50.320
our most crucial technologies.
00:05:50.320 --> 00:05:55.120
Could you talk us through a little bit how that came about and what its strengths and
00:05:55.120 --> 00:05:56.120
its weaknesses are?
00:05:56.120 --> 00:06:00.720
Well, I think you're right to illustrate the importance of the supply chain because Taiwan
00:06:00.720 --> 00:06:05.720
is the manufacturer of all these critical chips, but they can't do it alone.
00:06:05.720 --> 00:06:10.400
And in fact, many of the chips that Taiwan manufactures are actually designed by companies
00:06:10.400 --> 00:06:14.520
outside of Taiwan, Apple, NVIDIA, AMD, Qualcomm.
00:06:14.520 --> 00:06:19.800
It's actually largely US firms that are designing chips and then sending those designs in Taiwan
00:06:19.800 --> 00:06:20.800
to get manufactured.
00:06:20.800 --> 00:06:27.880
But the manufacturing process itself that occurs in Taiwan depends on inputs from Europe,
00:06:27.880 --> 00:06:30.320
from Japan, and from the United States.
00:06:30.320 --> 00:06:34.960
So for example, the ultra precise tools that are used to make chips come from the Netherlands
00:06:34.960 --> 00:06:37.160
and Japan, as well as the US.
00:06:37.160 --> 00:06:42.800
The materials that are used in chip making also come from these countries.
00:06:42.800 --> 00:06:47.160
And so Taiwan is critically important, but on its own it can't do anything without its
00:06:47.160 --> 00:06:51.440
relations with US and European and Japanese firms.
00:06:51.440 --> 00:06:55.320
And so thinking about the entire supply chain is critically important.
00:06:55.320 --> 00:07:00.160
The good news is that most of this advanced semiconductor supply chain actually does happen
00:07:00.160 --> 00:07:03.960
in the US, in Europe, or in Japan.
00:07:03.960 --> 00:07:09.640
The unfortunate news is that there's a surprising amount of concentration in Taiwan, and it's
00:07:09.640 --> 00:07:14.840
become more concentrated in Taiwan just as the risk that I think China poses to Taiwan
00:07:14.840 --> 00:07:16.120
has grown.
00:07:16.120 --> 00:07:17.720
And that's the core of the problem.
00:07:17.720 --> 00:07:18.720
Right.
00:07:18.720 --> 00:07:23.640
And you say it's unfortunate, of course, the Taiwanese, some Taiwanese politicians, some
00:07:23.640 --> 00:07:26.480
Taiwanese people talk about this as their silicon shield.
00:07:26.480 --> 00:07:30.000
I think you're right to focus on the ways in which it is unfortunate.
00:07:30.000 --> 00:07:34.280
I don't think it's at all unfortunate that Taiwan has developed this critical industry.
00:07:34.280 --> 00:07:38.520
I think it's actually great for Taiwan, and in a lot of ways great for the world.
00:07:38.520 --> 00:07:44.120
The problem is that our ability to defend Taiwan has been deteriorating at the very same
00:07:44.120 --> 00:07:49.240
time as we become more reliant on Taiwan for the chips that power our entire technological
00:07:49.240 --> 00:07:50.400
and industrial base.
00:07:50.400 --> 00:07:56.640
And so if we had the ability to deter China at the level at which we could 30 years ago,
00:07:56.640 --> 00:07:57.640
this wouldn't be a problem.
00:07:57.640 --> 00:08:01.920
This would be nothing but a success story for Taiwan and for the rest of us who have benefited
00:08:01.920 --> 00:08:06.960
from the ultra precise, capable, and efficient chips that Taiwan can produce.
00:08:06.960 --> 00:08:12.040
The problem is that just as the geopolitical picture around Taiwan has darkened as Western
00:08:12.040 --> 00:08:17.000
and US deterrence capabilities have, I think, eroded, we've also become more reliant at
00:08:17.000 --> 00:08:18.000
the same time.
00:08:18.000 --> 00:08:22.960
And so it's really the combination of this dependence plus critically the, I think,
00:08:22.960 --> 00:08:27.320
growing Chinese threat to Taiwan that makes this such a problematic issue.
00:08:27.320 --> 00:08:33.200
So then does our thinking around geopolitics need to change then?
00:08:33.200 --> 00:08:37.680
Does our mindset need to change in order to have a blueprint for how we are to increase
00:08:37.680 --> 00:08:39.480
those deterrence capabilities again?
00:08:39.480 --> 00:08:44.520
So the quote again that we started the show with chips are the new oil is East Asia now
00:08:44.520 --> 00:08:46.760
the new Middle East.
00:08:46.760 --> 00:08:51.760
And how does that concentration that you've talked about of the semiconductor industry,
00:08:51.760 --> 00:08:59.160
the geography of its knowhow and manufacturing, redraw potential geopolitical hotspots?
00:08:59.160 --> 00:09:06.280
And how do our assumptions need to change about what the conflict of the future might
00:09:06.280 --> 00:09:07.280
be?
00:09:07.280 --> 00:09:08.280
Were that ever to come to pass?
00:09:08.280 --> 00:09:10.160
Chris takes a moment to think that one through.
00:09:10.160 --> 00:09:14.320
I'd just like to direct all of our listeners inquiries about that comparison to Aaron via
00:09:14.320 --> 00:09:15.320
Twitter.
00:09:15.320 --> 00:09:16.320
Thank you.
00:09:16.320 --> 00:09:18.000
I think that is the right comparison to make.
00:09:18.000 --> 00:09:24.880
We've all spent the last half century learning about the challenges posed by advanced economies
00:09:24.880 --> 00:09:27.360
reliance on Middle Eastern oil.
00:09:27.360 --> 00:09:32.640
And that's posed a whole set of geopolitical challenges, which have been very complex and
00:09:32.640 --> 00:09:38.200
we've spent many decades debating how best to mitigate and balance these risks.
00:09:38.200 --> 00:09:41.800
If you look at the chip industry, it's even more concentrated than oil production.
00:09:41.800 --> 00:09:45.840
Saudi Arabia produces 10 or 15% of the world's oil depending on the year.
00:09:45.840 --> 00:09:48.480
Taiwan produces 90% of the most advanced processors.
00:09:48.480 --> 00:09:51.000
So there's even more concentration here.
00:09:51.000 --> 00:09:55.000
And there's more concentration because the chip making process is harder than oil drilling.
00:09:55.000 --> 00:09:56.800
Lots of countries can drill for oil.
00:09:56.800 --> 00:10:01.840
There's some technological complexity, especially when it comes to fracking or horizontal drilling
00:10:01.840 --> 00:10:08.040
or deep offshore, but the level of complexity is substantially less than in chip making.
00:10:08.040 --> 00:10:13.000
And so I think that does mean that we've got to begin to realize that when we're thinking
00:10:13.000 --> 00:10:19.040
about security in East Asia, it's not just about security dynamics that matter.
00:10:19.040 --> 00:10:23.840
Certainly Taiwan security is important in its own right and it has huge implications for
00:10:23.840 --> 00:10:27.480
the rest of the world order, but it also has huge economic implications too.
00:10:27.480 --> 00:10:32.360
And I think you can't understand potential crises that might emerge in East Asia without
00:10:32.360 --> 00:10:37.160
thinking about the economics alongside the security dynamics, just like you can't understand
00:10:37.160 --> 00:10:42.200
any Middle Eastern crisis without thinking about how every government is trying to reason
00:10:42.200 --> 00:10:46.440
through how oil price dynamics will shape other countries' reaction functions.
00:10:46.440 --> 00:10:50.920
The same thing is true for chips and electronic supply chains in East Asia.
00:10:50.920 --> 00:10:51.920
Right.
00:10:51.920 --> 00:10:55.960
And just to give people a sense of this, of both the scale and the complexity involved,
00:10:55.960 --> 00:11:00.720
you mentioned before the lithography machines that are used for producing these high-end
00:11:00.720 --> 00:11:01.720
chips.
00:11:01.720 --> 00:11:05.640
And Europe obviously has one of the global champions of that industry, ASML, the Dutch
00:11:05.640 --> 00:11:11.840
conglomerate, which formed by basically buying up their entire supply chain in order to control
00:11:11.840 --> 00:11:16.200
production processes to the level that was needed in order to make their world-beating
00:11:16.200 --> 00:11:17.440
technology.
00:11:17.440 --> 00:11:21.840
And if you look at where they source from over the world, it's really a sort of hyper-globalization
00:11:21.840 --> 00:11:27.640
story that's been told, and it relies on that flat world of the 1990s, which then came to
00:11:27.640 --> 00:11:33.360
pass through corporate structuring in the 2000s and the 2010s.
00:11:33.360 --> 00:11:37.200
And this worked, and I think this is what your book brings out quite brilliantly as well,
00:11:37.200 --> 00:11:38.680
this worked because of the scale involved.
00:11:38.680 --> 00:11:41.960
If you think about the distances, the things involved have to travel.
00:11:41.960 --> 00:11:45.480
It sometimes blows the mind to think this is a more efficient or cost-effective way
00:11:45.480 --> 00:11:46.480
of doing things.
00:11:46.480 --> 00:11:51.040
But I think one anecdote from your book that really brings this home, which I'm going to
00:11:51.040 --> 00:11:57.760
just quote Sue to read is, to understand how much is involved in each one of these chips.
00:11:57.760 --> 00:12:06.160
So you note that Apple sold 100 million iPhone 12s, each powered by this A14 processor chip,
00:12:06.160 --> 00:12:11.640
which has 11.8 billion transistors carved into its silicon.
00:12:11.640 --> 00:12:14.160
And I thought, yeah, this is amazing the way you brought this out, you say.
00:12:14.160 --> 00:12:19.760
So in a matter of months, for just one of the dozen chips in an iPhone, TSMC in Taiwan
00:12:19.760 --> 00:12:24.400
fabricated over one quintillion transistors.
00:12:24.400 --> 00:12:27.440
And that's a number, as you say, with 18 zeros behind it.
00:12:27.440 --> 00:12:31.960
Now to put that in context, last year, the chip industry produced more transistors than
00:12:31.960 --> 00:12:38.520
the combined quantity of all goods produced by all other companies in all other industries
00:12:38.520 --> 00:12:41.040
in all of human history.
00:12:41.040 --> 00:12:45.080
And that is the scale that makes this global efficiency work.
00:12:45.080 --> 00:12:46.080
And that's what we're dealing with.
00:12:46.080 --> 00:12:49.640
So unscrambling those eggs is going to be pretty tricky, isn't it, Chris?
00:12:49.640 --> 00:12:50.640
That's right.
00:12:50.640 --> 00:12:54.040
And the goal, I think, is not to completely unscramble those eggs.
00:12:54.040 --> 00:13:00.120
We've got to keep the scale dynamics in place, which is why when I hear political leaders
00:13:00.120 --> 00:13:03.760
occasionally talk about self-sufficiency, I get very nervous.
00:13:03.760 --> 00:13:09.720
Because self-sufficiency is not even a realistic goal, but trying to achieve it, I think, would
00:13:09.720 --> 00:13:15.600
have vast and unfortunate consequences in terms of creating inefficiencies that would
00:13:15.600 --> 00:13:18.560
slow technological progress and drive up costs.
00:13:18.560 --> 00:13:23.160
The challenge is that we, of course, can't keep, I don't think, technology supply chains
00:13:23.160 --> 00:13:27.520
structured in the way they are, both in terms of their entanglement with China and in terms
00:13:27.520 --> 00:13:32.200
of the points of concentration they have right outside of China.
00:13:32.200 --> 00:13:36.040
And so that is the balance that I think Western leaders are trying to strike right now.
00:13:36.040 --> 00:13:42.200
Japanese leaders, European leaders, US leaders, is to do enough restructuring of supply chains
00:13:42.200 --> 00:13:47.720
to gain more margin of security, but to not impose too great of cost on the industry that
00:13:47.720 --> 00:13:50.520
would have the effect of slowing technological progress.
00:13:50.520 --> 00:13:58.720
And so to me, that's why we've got to have as much coordination as possible between G7
00:13:58.720 --> 00:14:03.800
countries and between close allies in making sure that we've got a shared space for supply
00:14:03.800 --> 00:14:08.600
chains to develop even further, to have this division of labor between countries.
00:14:08.600 --> 00:14:13.480
But we've also got limits when it comes to trade with countries that we can't trust.
00:14:13.480 --> 00:14:16.680
And that's the balance that I think we are trying to strike right now.
00:14:16.680 --> 00:14:20.880
And that comes actually to something that we've discussed on Berlin's side out several times,
00:14:20.880 --> 00:14:25.320
the national security premium, or indeed, friend-shoring.
00:14:25.320 --> 00:14:31.080
And when we talk about China, this word comes up a lot, the idea that we should trade more
00:14:31.080 --> 00:14:35.920
with our friends and also have supply chains that are more integrated with our friends.
00:14:35.920 --> 00:14:40.040
That, of course, brings up the question, who is a friend?
00:14:40.040 --> 00:14:43.400
And where do you manufacture, for example?
00:14:43.400 --> 00:14:51.160
But the other example that comes to mind here, too, is ASML in the Netherlands.
00:14:51.160 --> 00:14:57.040
I mean, they were none too happy about the Dutch government restricting their sales into
00:14:57.040 --> 00:14:58.760
the Chinese market.
00:14:58.760 --> 00:15:02.680
And first, they couldn't sell their most advanced lithography machines, and they couldn't even
00:15:02.680 --> 00:15:04.720
service machines that they had already sold in China.
00:15:04.720 --> 00:15:07.280
And then they accused the US of putting its finger on the scale.
00:15:07.280 --> 00:15:14.360
But how far apart are we really between the US and Europe, for example, in terms of how
00:15:14.360 --> 00:15:20.960
we actually see friend-shoring when it comes to this industry in particular?
00:15:20.960 --> 00:15:22.960
Where do you think the lay of the land is?
00:15:22.960 --> 00:15:28.800
Well, I think there's a debate within Europe, within the United States, and also within Japan,
00:15:28.800 --> 00:15:35.720
which is worth considering as well, between business and government as to what level of
00:15:35.720 --> 00:15:37.040
friend-shoring makes sense.
00:15:37.040 --> 00:15:40.800
I think business naturally wants to focus on efficiency.
00:15:40.800 --> 00:15:45.360
Businesses naturally don't want to restructure supply chains unless they absolutely have to.
00:15:45.360 --> 00:15:48.840
And so the same debate that's happening between ASML and the Dutch government is happening
00:15:48.840 --> 00:15:53.640
between US chip firms and the US government and Japanese firms and the Japanese government.
00:15:53.640 --> 00:15:55.200
So that's understandable.
00:15:55.200 --> 00:15:58.120
It's good that businesses are making their voice heard to some extent.
00:15:58.120 --> 00:16:01.360
Governments need to consider their interests.
00:16:01.360 --> 00:16:07.000
But I think in all three of these regions, US, Japan, Europe, governments are much more
00:16:07.000 --> 00:16:11.880
concerned than they were a couple of years ago about becoming too reliant on China or
00:16:11.880 --> 00:16:17.880
areas that are exposed to China risk, partly because of the way China's changed in the
00:16:17.880 --> 00:16:24.480
last couple of years, partly because the Russian invasion of Ukraine has made very real, I
00:16:24.480 --> 00:16:30.360
think, the reality that authoritarian governments can do very dangerous and seemingly risky
00:16:30.360 --> 00:16:35.440
things, and partly because the pandemic, although it wasn't related to geopolitical
00:16:35.440 --> 00:16:40.160
shocks, did illustrate the extent to which supply chains for all types of goods, not
00:16:40.160 --> 00:16:45.560
just advanced chips, have become so complex in ways that governments and even companies
00:16:45.560 --> 00:16:46.560
only barely understand.
00:16:46.560 --> 00:16:52.240
And so I think the governments have across the board been more willing to push towards
00:16:52.240 --> 00:16:53.640
French shoring as a result.
00:16:53.640 --> 00:16:57.800
I think in Europe, there's differences between countries, of course.
00:16:57.800 --> 00:17:04.280
I think the Dutch are actually on one side of the spectrum in my experience and being
00:17:04.280 --> 00:17:10.320
more skeptical of deep economic interconnections with China, more concerned about China in
00:17:10.320 --> 00:17:11.640
a political sense.
00:17:11.640 --> 00:17:15.960
I think if you look at some southern European countries and also Germany, you get a different
00:17:15.960 --> 00:17:16.960
view.
00:17:16.960 --> 00:17:22.400
I think Chancellor Scholz's visit to China not too long ago was a very different type
00:17:22.400 --> 00:17:27.120
of visit than you saw when the Dutch Prime Minister visited earlier this year.
00:17:27.120 --> 00:17:29.800
And so I think that does certainly stand out to me.
00:17:29.800 --> 00:17:31.920
Indeed, a different approach is taken.
00:17:31.920 --> 00:17:37.560
And you begin the book talking actually about the chip choke and how the US in particular
00:17:37.560 --> 00:17:41.640
was able to put the squeeze on China to cut off their new oil.
00:17:41.640 --> 00:17:46.240
Could you tell us a bit more about that and what levers we actually have that we can pull?
00:17:46.240 --> 00:17:52.320
The US has been trying to pursue technology restrictions in two different spheres.
00:17:52.320 --> 00:17:57.240
One is with regard to advanced AI chips, the types of chips that are used to train open
00:17:57.240 --> 00:18:02.080
AI systems or anthropic systems, the most complex AI models.
00:18:02.080 --> 00:18:06.280
And today these chips are produced largely by a couple of US firms and manufactured largely
00:18:06.280 --> 00:18:07.280
in Taiwan.
00:18:07.280 --> 00:18:12.360
And now since 2022, they've been illegal to transfer to China, at least referring to
00:18:12.360 --> 00:18:14.360
the most advanced types of these chips.
00:18:14.360 --> 00:18:18.560
The second type of restriction that the US has been imposing alongside the Dutch government
00:18:18.560 --> 00:18:23.280
and alongside the Japanese government relates to chip making tools like the lithography
00:18:23.280 --> 00:18:25.280
tools that ASML produces.
00:18:25.280 --> 00:18:29.680
And the rationale here is that if China can buy the most advanced chip making tools, it
00:18:29.680 --> 00:18:34.760
can then make the most advanced chips itself and then use them for whatever purposes it
00:18:34.760 --> 00:18:35.760
likes.
00:18:35.760 --> 00:18:43.760
And so the US strategy ultimately is described as being about Chinese military capabilities.
00:18:43.760 --> 00:18:45.760
And that's true to an extent.
00:18:45.760 --> 00:18:50.320
It's certainly the case that China will use whatever AI chips it has for military as
00:18:50.320 --> 00:18:52.080
well as civilian capabilities.
00:18:52.080 --> 00:18:57.320
And I think to really understand the US strategy, it's about something deeper, which is which
00:18:57.320 --> 00:19:02.160
block is going to create the infrastructure on which the next generation of technology
00:19:02.160 --> 00:19:03.600
is built.
00:19:03.600 --> 00:19:08.640
The West has benefited from the fact that over the past half century, its companies have
00:19:08.640 --> 00:19:13.320
largely built this critical technology infrastructure with one glaring exception, and that's Huawei
00:19:13.320 --> 00:19:15.080
and the telecom space.
00:19:15.080 --> 00:19:20.960
Now we're building a new generation of infrastructure for a world that will see AI be a primary driver
00:19:20.960 --> 00:19:22.800
of productivity increases.
00:19:22.800 --> 00:19:27.080
And so the question of, well, who builds this infrastructure is, I think, not just a technical
00:19:27.080 --> 00:19:29.120
question nor an economic question.
00:19:29.120 --> 00:19:31.680
It's fundamentally a political question.
00:19:31.680 --> 00:19:36.560
The US is focused on making sure that it's Western companies that are doing this.
00:19:36.560 --> 00:19:38.200
And it's worth underscoring.
00:19:38.200 --> 00:19:41.880
It's not just US firms that are part of this infrastructure.
00:19:41.880 --> 00:19:43.680
It's the Taiwanese who actually make the chips.
00:19:43.680 --> 00:19:46.840
It's the Dutch that make the machines that make the chips.
00:19:46.840 --> 00:19:53.440
And so I think seeing this as a broader Western strategy is actually the accurate way to look
00:19:53.440 --> 00:19:59.400
at what the US has been trying to accomplish and pushing its companies and also its friends
00:19:59.400 --> 00:20:00.960
towards a more restrictive strategy.
00:20:00.960 --> 00:20:01.960
Absolutely.
00:20:01.960 --> 00:20:04.360
And Chris, for how long do we have the edge in the current system?
00:20:04.360 --> 00:20:08.240
How far away is China from actually achieving Silicon sovereignty?
00:20:08.240 --> 00:20:12.800
I would say first off that the restrictions that are currently in place have already slowed
00:20:12.800 --> 00:20:14.680
China's progress.
00:20:14.680 --> 00:20:20.600
It's now been six years since Dutch regulations have prohibited Chinese firms from acquiring
00:20:20.600 --> 00:20:25.200
the most advanced lithography tools that are used to make the most advanced chips.
00:20:25.200 --> 00:20:29.880
And if China had been able to buy those tools six years ago, it would be in a much, much
00:20:29.880 --> 00:20:32.360
stronger position than it is today.
00:20:32.360 --> 00:20:34.360
Because it can't buy those tools.
00:20:34.360 --> 00:20:39.600
China's leading chipmaker is around five years behind the TSMC of Taiwan in terms of its
00:20:39.600 --> 00:20:42.160
ability to manufacture advanced chips.
00:20:42.160 --> 00:20:45.920
And I think the Western goal is to keep that gap in place.
00:20:45.920 --> 00:20:51.800
And five years might not sound like much because the quality of chips doubles every two years.
00:20:51.800 --> 00:20:56.280
Five years is two and a half generations of differential ballpark terms.
00:20:56.280 --> 00:21:02.160
And so the US is trying, alongside the Dutch and the Japanese, to hold back China's progress.
00:21:02.160 --> 00:21:06.360
But I think it's equally important, probably more important to focus on the fact that most
00:21:06.360 --> 00:21:11.440
Western governments are also doubling down on their efforts to spur technological progress.
00:21:11.440 --> 00:21:15.160
Because ultimately, you're not going to win if you only hold China back.
00:21:15.160 --> 00:21:17.920
More important than that even is racing more rapidly forward yourself.
00:21:17.920 --> 00:21:22.640
And that's what efforts like the EU Chips Act and the US Chips Act are partly intended
00:21:22.640 --> 00:21:23.640
to achieve.
00:21:23.640 --> 00:21:25.240
This is, in some ways, nothing new.
00:21:25.240 --> 00:21:28.320
And I think that's something that the book brings out very well as well.
00:21:28.320 --> 00:21:31.520
Is that technology gap that existed in the Cold War time?
00:21:31.520 --> 00:21:36.440
I mean, you very, very well illustrate how the Soviet strategy of copy it, steal the
00:21:36.440 --> 00:21:40.080
technology and copy it, actually condemned them to relative backwardness throughout the
00:21:40.080 --> 00:21:44.600
whole Cold War and kept them five years behind, roughly speaking, which, as you say, is a
00:21:44.600 --> 00:21:47.320
world away in terms of the tech that was there.
00:21:47.320 --> 00:21:51.480
But of course, also, that was a time when the US was trying to maximize its competitive
00:21:51.480 --> 00:21:56.000
advantage economically, as well as in the military race against the Soviets, as well
00:21:56.000 --> 00:21:59.200
as in that geopolitical competition.
00:21:59.200 --> 00:22:01.920
And it came together with questions of, again, who are our friends?
00:22:01.920 --> 00:22:06.680
Because of course, Silicon Valley only really took off when they shifted production to Japan.
00:22:06.680 --> 00:22:07.680
That's right.
00:22:07.680 --> 00:22:13.280
And it is, I think, important to note that the chip industry has been internationalized
00:22:13.280 --> 00:22:14.560
since essentially day one.
00:22:14.560 --> 00:22:17.720
The first chips were invented in 1958-59.
00:22:17.720 --> 00:22:23.680
The first offshore assembly facility was set up in Hong Kong in 1963.
00:22:23.680 --> 00:22:30.360
So from almost day one, there have been trans-Pacific ties that have been critical to the structure
00:22:30.360 --> 00:22:31.520
of the industry.
00:22:31.520 --> 00:22:34.320
And Europe's played a very big role from the earliest stage.
00:22:34.320 --> 00:22:37.640
And as we noted before, that's been key to technological progress.
00:22:37.640 --> 00:22:42.840
Having this division of labor between different companies and different countries.
00:22:42.840 --> 00:22:47.520
And it's meant that we've always had to think about chip strategy in the context of other
00:22:47.520 --> 00:22:49.560
markets and other friends.
00:22:49.560 --> 00:22:54.800
And this worked out relatively straightforwardly until pretty recently, because there was commercial
00:22:54.800 --> 00:23:00.320
competition between US firms, Japanese firms, and German firms, and Korean firms.
00:23:00.320 --> 00:23:04.640
But most of the technology leaders were in friendly countries.
00:23:04.640 --> 00:23:08.600
The difference now over the past couple of years is that Chinese firms have become real
00:23:08.600 --> 00:23:10.360
technology challengers.
00:23:10.360 --> 00:23:13.880
And they've become technology challengers at the exact same time as China's government
00:23:13.880 --> 00:23:20.360
has become a systemic challenger to Western leadership in the world and the Western dominated
00:23:20.360 --> 00:23:21.360
world order.
00:23:21.360 --> 00:23:24.720
And so you can't see this today simply as commercial competition.
00:23:24.720 --> 00:23:29.040
And I often hear people describe what's happening as a trade war.
00:23:29.040 --> 00:23:32.400
And that completely misunderstands the stakes.
00:23:32.400 --> 00:23:36.720
There's no strategic interest of the United States or of Europe in promoting this company
00:23:36.720 --> 00:23:40.960
or that company, except that these companies are critical in laying down the infrastructure
00:23:40.960 --> 00:23:43.840
on which the rest of economy and society depend.
00:23:43.840 --> 00:23:47.400
And so it's not just about corporate earnings in this quarter or that quarter.
00:23:47.400 --> 00:23:50.000
It's not about the trade deficit or the trade surplus.
00:23:50.000 --> 00:23:53.160
It's about who gets to write the rules of the next generation of technology.
00:23:53.160 --> 00:23:57.920
And I want that to be companies that are based in countries with similar sets of values and
00:23:57.920 --> 00:23:58.920
not China.
00:23:58.920 --> 00:23:59.920
Right.
00:23:59.920 --> 00:24:06.080
And you mentioned just now sort of something that I almost see is the value of friendship
00:24:06.080 --> 00:24:12.600
between Western countries and technological leaders in those countries, this kind of
00:24:12.600 --> 00:24:16.440
exchange between the democratic world when it comes to this particular industry.
00:24:16.440 --> 00:24:21.360
And that seems to give us an edge in this particular case, at least so far.
00:24:21.360 --> 00:24:28.000
You have argued before that Chinese chipmanking is becoming more Soviet in a way and that its
00:24:28.000 --> 00:24:34.480
own authoritarian politics in contrast to cooperation could actually hamper its ability
00:24:34.480 --> 00:24:37.760
to catch up to the know-how in democratic countries.
00:24:37.760 --> 00:24:40.000
And so what makes you say this?
00:24:40.000 --> 00:24:44.680
Does liberal democracy have an advantage here when it comes to staying ahead on chips?
00:24:44.680 --> 00:24:49.840
We should differentiate somewhat between liberal democracies and market economies.
00:24:49.840 --> 00:24:55.840
I think market economies probably are the key driver of advantage rather than liberal
00:24:55.840 --> 00:24:57.120
democracy.
00:24:57.120 --> 00:25:03.200
But I think democracy does explain why property rights are more durable in the West than they
00:25:03.200 --> 00:25:04.200
have been in China.
00:25:04.200 --> 00:25:08.720
And if you look at one of the key challenges to China's tech sector over the last five
00:25:08.720 --> 00:25:14.760
years has been the politically driven crackdown on Chinese tech companies, whereby the fact
00:25:14.760 --> 00:25:19.840
that China doesn't have the rule of law has in fact undermined property rights, discouraged
00:25:19.840 --> 00:25:21.480
investment in China's own tech sector.
00:25:21.480 --> 00:25:28.440
So there is a linkage between the democratic aspect and the actual effective functioning
00:25:28.440 --> 00:25:30.040
of a free market.
00:25:30.040 --> 00:25:36.160
I think, however, we shouldn't rest on that as the key source of our advantage.
00:25:36.160 --> 00:25:42.040
I think it's probably true that over the long run, if you had two equal systems in terms
00:25:42.040 --> 00:25:47.520
of capabilities, the liberal free market economy went out over the statist economy without
00:25:47.520 --> 00:25:48.880
the rule of law.
00:25:48.880 --> 00:25:52.760
And I think we're in the midst of a really tough competition regardless.
00:25:52.760 --> 00:25:58.000
And that's even considering the fact that US plus Europe plus Japan GDP is substantially
00:25:58.000 --> 00:26:00.400
larger than China's.
00:26:00.400 --> 00:26:03.520
The reality is China's got very successful firms.
00:26:03.520 --> 00:26:05.480
It's got very impressive engineers.
00:26:05.480 --> 00:26:07.560
It's got extraordinary focus on this issue.
00:26:07.560 --> 00:26:12.560
And so pretending that we can do nothing because our system is better, I think is a recipe for
00:26:12.560 --> 00:26:13.560
failure.
00:26:13.560 --> 00:26:15.760
I think if you want to win the competition, you've got to compete.
00:26:15.760 --> 00:26:16.760
Right.
00:26:16.760 --> 00:26:20.280
And I think, again, this is a recurring debate that comes up through the book is how to actually
00:26:20.280 --> 00:26:24.720
harness those different sources of our power, how to align them in ways that are advantageous
00:26:24.720 --> 00:26:29.200
to us rather than our enemies to really make good on that promise that is there, that we
00:26:29.200 --> 00:26:32.850
We can't just rest on our laurels, but we actually have to activate this advantage.
00:26:32.850 --> 00:26:37.250
I'd like to hear a bit more from you, Christopher, how that was done at various stages during
00:26:37.250 --> 00:26:38.750
in the past.
00:26:38.750 --> 00:26:43.830
Be that through Bill Perry's offset strategy and actually sort of giving a boost to the
00:26:43.830 --> 00:26:50.330
Silicon industry there, or earlier in the very beginning of the semiconductor industry.
00:26:50.330 --> 00:26:55.690
How those different aspects of US society, be that education, science, technology, military,
00:26:55.690 --> 00:26:59.890
and then the market came to actually give the US the advantage in the first place, the
00:26:59.890 --> 00:27:01.490
US and the West, I should say.
00:27:01.490 --> 00:27:07.650
I think I would start by looking at the question of who funds new innovations in this sector.
00:27:07.650 --> 00:27:13.050
We're used to, I think, in the software space, having big venture capital funds, funding
00:27:13.050 --> 00:27:18.730
new companies that then go on to create apps or social media platforms.
00:27:18.730 --> 00:27:24.330
But in the semiconductor space and the broader tech hardware space, what you find is that
00:27:24.330 --> 00:27:29.810
governments are much more important in terms of putting up funds because the capital requirements
00:27:29.810 --> 00:27:34.210
are larger, the technology is more complex, there's a longer time from innovation to reaching
00:27:34.210 --> 00:27:35.210
a market.
00:27:35.210 --> 00:27:40.770
So, dating back all the way to the start of the chip industry, it was NASA, for example,
00:27:40.770 --> 00:27:45.010
the US Space Agency, or it was the Department of Defense that were funding the core technology
00:27:45.010 --> 00:27:46.570
development.
00:27:46.570 --> 00:27:51.690
That was critical to making it possible to improve technology enough to the point where
00:27:51.690 --> 00:27:55.810
companies could pick it up and then introduce it to a broad civilian market.
00:27:55.810 --> 00:28:02.570
And I think to some extent over the past couple of decades, we have collectively in the West
00:28:02.570 --> 00:28:06.730
under-invested in this type of technological development, in part because a lot of the
00:28:06.730 --> 00:28:11.050
reason we were doing more of it was due to Cold War competition.
00:28:11.050 --> 00:28:15.810
It was Cold War competition that induced Western governments to put money in this sphere and
00:28:15.810 --> 00:28:20.730
when geopolitical competition seemed to temporarily go away in the '90s and 2000s.
00:28:20.730 --> 00:28:22.930
So to some of the funding dried up.
00:28:22.930 --> 00:28:26.530
I think that is starting to change and we've got to make sure that persists because that
00:28:26.530 --> 00:28:31.050
funding is really critical for early stage technological development.
00:28:31.050 --> 00:28:39.250
The second thing is there's long been a sort of implicit assumption in Western circles
00:28:39.250 --> 00:28:44.530
that if we pursue free markets, everyone else must be pursuing them too.
00:28:44.530 --> 00:28:50.530
And so I often, for example, in the space of the chip industry, see concerns about Western
00:28:50.530 --> 00:28:56.410
governments starting a subsidy race by supporting chip investments over the past couple of years
00:28:56.410 --> 00:28:58.850
via the EU CHIPS Act and the US CHIPS Act.
00:28:58.850 --> 00:29:02.250
And of course, you would only think that the West was starting a subsidy race if you knew
00:29:02.250 --> 00:29:05.090
nothing about China's chip policies.
00:29:05.090 --> 00:29:07.810
And in fact, there's been a subsidy race ongoing for a long time.
00:29:07.810 --> 00:29:12.050
We've just only decided to compete in it in the last couple of years.
00:29:12.050 --> 00:29:15.210
And so there too, I think you've got to look very carefully at the roles governments are
00:29:15.210 --> 00:29:16.210
playing.
00:29:16.210 --> 00:29:20.130
Would it be great if we had a level playing field in the chip industry and China was playing
00:29:20.130 --> 00:29:21.130
with the same rules?
00:29:21.130 --> 00:29:22.130
That would be great.
00:29:22.130 --> 00:29:23.130
But it's not the world we live in.
00:29:23.130 --> 00:29:24.690
And so I think we've got to be realistic.
00:29:24.690 --> 00:29:29.810
If China is going to subsidize to an extraordinary degree, chip facilities will understandably
00:29:29.810 --> 00:29:31.650
move to high subsidy locations.
00:29:31.650 --> 00:29:36.010
And so that's the second sphere where government policy really does play an absolutely critical
00:29:36.010 --> 00:29:37.010
role.
00:29:37.010 --> 00:29:40.850
And understanding the nature of that geopolitical challenge as well as the playing field, as
00:29:40.850 --> 00:29:45.690
you put it, seems to be key in being able to encapsulate that in a strategic thinking
00:29:45.690 --> 00:29:49.890
that can prioritize these kind of connections between those different sectors that really
00:29:49.890 --> 00:29:55.410
gave the advantage to the semiconductor industry in the West during the Cold War is our challenge
00:29:55.410 --> 00:29:58.050
renewed in different circumstances, it would seem.
00:29:58.050 --> 00:30:03.250
Now, I did mention this term offset, and that's been a consistent theme of using technology
00:30:03.250 --> 00:30:07.250
to offset other disadvantages or the advantages of others.
00:30:07.250 --> 00:30:13.290
And in the 1970s, that was the feeling that the need in the West to offset the growing
00:30:13.290 --> 00:30:17.810
nuclear forces of the Soviet Union, but also there are their superior conventional capabilities,
00:30:17.810 --> 00:30:22.690
they could outmatch or overmatch NATO's conventional capabilities.
00:30:22.690 --> 00:30:27.090
We heard the term offset again used in the Obama administration.
00:30:27.090 --> 00:30:31.090
Was that actually successful in offsetting the advantages that others were building,
00:30:31.090 --> 00:30:33.130
or do we need a new offset again now?
00:30:33.130 --> 00:30:38.890
Well, I think we're actually still in the process of developing the offset that was
00:30:38.890 --> 00:30:41.610
set off in the late Obama administration.
00:30:41.610 --> 00:30:48.570
I think it's certainly true that Western defense establishments partly because of the perceived
00:30:48.570 --> 00:30:53.290
lack of geopolitical competition, partly because of their focus on counterterrorism in the
00:30:53.290 --> 00:30:59.090
2000s have not been thinking hard enough until recently about the ways that new types of
00:30:59.090 --> 00:31:01.850
technology will change warfare.
00:31:01.850 --> 00:31:07.890
I think that has changed dramatically, especially given the shock of the Russia-Ukraine war.
00:31:07.890 --> 00:31:13.610
But in order to succeed in updating the way Western militaries fight to next generation
00:31:13.610 --> 00:31:18.810
technological capabilities, we need sustained investments and sustained focus over a long
00:31:18.810 --> 00:31:19.810
period of time.
00:31:19.810 --> 00:31:24.130
And I think on the investment side, we're starting to see this happening in Europe and
00:31:24.130 --> 00:31:25.130
the US.
00:31:25.130 --> 00:31:29.850
In terms of focus, I think one of the key challenges is that it's one thing to spend more money,
00:31:29.850 --> 00:31:33.410
it's another thing to convince the bureaucracy to do things in different ways.
00:31:33.410 --> 00:31:38.650
And militaries are very big bureaucracies and it takes a lot of sustained political focus
00:31:38.650 --> 00:31:41.370
to get them to shift their priorities.
00:31:41.370 --> 00:31:45.810
And that's where I think you need very senior political leadership that are making sure that
00:31:45.810 --> 00:31:49.770
militaries aren't taking their new resources and buying old systems, but rather taking
00:31:49.770 --> 00:31:55.010
their new resources and buying new systems that will be useful in new ways.
00:31:55.010 --> 00:31:58.370
And we see this to some extent, but I think there's a lot of work to be done.
00:31:58.370 --> 00:32:04.850
And it's particularly important that I think we're very careful in any limitations we impose
00:32:04.850 --> 00:32:07.690
on the use of new technologies.
00:32:07.690 --> 00:32:12.930
Western publics are understandably nervous when they hear about space as a new domain
00:32:12.930 --> 00:32:18.650
for military competition or semi-autonomous systems being deployed in the military sphere.
00:32:18.650 --> 00:32:24.050
But the reality is, these are the places where new technologies are going to be most impactful
00:32:24.050 --> 00:32:28.770
and our adversaries are going to not be constrained by any moral considerations that we might be
00:32:28.770 --> 00:32:29.770
buying.
00:32:29.770 --> 00:32:35.410
So we must be careful, but we also, I think, must not be overly constrained and prevent
00:32:35.410 --> 00:32:39.210
our militaries from developing the capabilities that they ultimately need to defend us.
00:32:39.210 --> 00:32:40.210
This is an interesting point.
00:32:40.210 --> 00:32:44.530
The British architectural theorist Jonathan Meads once said that necessity may be the
00:32:44.530 --> 00:32:47.370
mother of invention, but war is its birth mother.
00:32:47.370 --> 00:32:51.010
And actually understanding where this is going at the moment in Ukraine and the lessons that
00:32:51.010 --> 00:32:54.450
are being learned is key if we are to retain an advantage.
00:32:54.450 --> 00:32:57.770
And there are German companies working on this, there are others working on this.
00:32:57.770 --> 00:33:03.010
But as we know, the German Titan vendor has been certainly slow to embrace the new.
00:33:03.010 --> 00:33:07.370
The main spending on defense is on an old shopping list already from the mid-2010s.
00:33:07.370 --> 00:33:12.290
It's not necessarily on the new kind of shaping technologies that we would need to maintain
00:33:12.290 --> 00:33:13.690
an edge.
00:33:13.690 --> 00:33:17.970
And Aaron, that does rather raise the question of what's expected of allies in this regard,
00:33:17.970 --> 00:33:18.970
doesn't it?
00:33:18.970 --> 00:33:19.970
Absolutely.
00:33:19.970 --> 00:33:28.530
And we've spoken before about the Gulf that exists between the US and Europe on even de-risking
00:33:28.530 --> 00:33:35.050
from China, how exactly, you know, and more American voices are in favor of sort of a
00:33:35.050 --> 00:33:42.010
full decoupling or towards a full decoupling, we should say, simply because what that dependence
00:33:42.010 --> 00:33:49.970
could do to us as the West should, that Chinese actually take steps to invade Taiwan for potentially
00:33:49.970 --> 00:33:54.330
the reason of taking control of something like a chip supply chain.
00:33:54.330 --> 00:33:59.570
That is a new incentive that goes on top of its territorial claim as we were discussing
00:33:59.570 --> 00:34:00.610
at the beginning of the show.
00:34:00.610 --> 00:34:07.650
But if the rubber does really meet the road on this, if an American warship is sunk by
00:34:07.650 --> 00:34:13.450
the Chinese and the Taiwan Strait, if China starts making moves to actually invade Taiwan,
00:34:13.450 --> 00:34:19.450
what do you think the US's actual expectations of Europe would be?
00:34:19.450 --> 00:34:26.130
I think one of the challenges of thinking through severe crises and shocks like this
00:34:26.130 --> 00:34:32.250
is that it's very difficult to predict how people's mental maps of what is possible changes.
00:34:32.250 --> 00:34:39.210
I think if you'd asked almost any European leader on February 20th, 2022, what their
00:34:39.210 --> 00:34:44.130
plausible responses would be to a full-scale Russian-Ukraine war, I don't think any of
00:34:44.130 --> 00:34:47.410
them would have predicted the types of steps they were willing to take two weeks later.
00:34:47.410 --> 00:34:52.450
And I think China-Taiwan dynamics have the same really fundamental uncertainties involved.
00:34:52.450 --> 00:34:57.490
I think there are people in Europe who think it might be possible for Europe to sort of
00:34:57.490 --> 00:35:01.730
stand aside in the case of a crisis like that.
00:35:01.730 --> 00:35:07.130
In my view, that would be impossible for political reasons, for economic reasons, for military
00:35:07.130 --> 00:35:08.850
reasons.
00:35:08.850 --> 00:35:13.890
Any sort of war between China and the United States would draw in basically every region
00:35:13.890 --> 00:35:16.290
of the world if they liked it or not.
00:35:16.290 --> 00:35:21.490
And I think as a result of that, it's in Europe's interest to start thinking seriously about
00:35:21.490 --> 00:35:29.490
how Europe would manage because neutrality would be impossible in a context in which economic
00:35:29.490 --> 00:35:34.290
supply chains were completely disrupted, in which there was a war that would be fought
00:35:34.290 --> 00:35:39.050
not just in one theater, but probably in multiple geographic theaters, and in which Europe would
00:35:39.050 --> 00:35:44.610
find itself damaged, even if it was not directly participating in a way that would push it
00:35:44.610 --> 00:35:46.850
towards participation.
00:35:46.850 --> 00:35:51.330
And so I think all that suggests that Europe does need to start thinking about this issue
00:35:51.330 --> 00:35:53.090
very carefully.
00:35:53.090 --> 00:35:59.470
And to think that you can just stand aside, I think Ms. Reads, both the scope of disruption
00:35:59.470 --> 00:36:04.090
in purely in economic terms that Europe would face, but also the extent that the US and
00:36:04.090 --> 00:36:08.650
China would both be pushing extraordinarily heavily on Europe to take steps.
00:36:08.650 --> 00:36:10.450
And Europe couldn't simply do nothing.
00:36:10.450 --> 00:36:15.770
You've spoken just now, made it an excellent point, I think, suddenly what was inconceivable
00:36:15.770 --> 00:36:18.210
for is now a very real.
00:36:18.210 --> 00:36:24.450
But the drawback of acting that way is, again, you act reactively, not proactively in any
00:36:24.450 --> 00:36:26.690
kind of strategic sense.
00:36:26.690 --> 00:36:32.190
And right now, evidence from Europe indicates that we are extremely unprepared for what
00:36:32.190 --> 00:36:38.550
might happen if tensions with China, over Taiwan, or anything else were to increase.
00:36:38.550 --> 00:36:44.710
So what kind of steps do you think are necessary to start considering then, so that we're not
00:36:44.710 --> 00:36:49.990
consistently caught on the back foot all the time, should something like this happen?
00:36:49.990 --> 00:36:55.150
I think that the first is for European political leaders and also business leaders to start
00:36:55.150 --> 00:36:56.850
doing their own scenarios.
00:36:56.850 --> 00:37:03.410
What would happen if, again, in the US, there's been a real mental change, mental shift over
00:37:03.410 --> 00:37:09.730
the past couple of years as both government think tanks and companies have started doing
00:37:09.730 --> 00:37:12.210
scenarios of crises in the Taiwan Straits.
00:37:12.210 --> 00:37:17.170
And that has sensitized, I think, a lot of people to just how dramatic the costs could
00:37:17.170 --> 00:37:20.930
be and just how far reaching the impact would be.
00:37:20.930 --> 00:37:24.010
And there certainly are people in Europe thinking through these issues.
00:37:24.010 --> 00:37:28.150
But my sense is that the number of European political leaders who have participated in
00:37:28.150 --> 00:37:32.990
a Taiwan Straits war game is substantially lower than the number of American or Japanese
00:37:32.990 --> 00:37:35.270
leaders who have done so.
00:37:35.270 --> 00:37:39.790
And once you start participating and thinking through the second and third order conflicts,
00:37:39.790 --> 00:37:43.310
you begin to realize that this issue ought to be higher up on your priority list because
00:37:43.310 --> 00:37:47.910
the magnitude of impact it would have on you is quite large.
00:37:47.910 --> 00:37:52.670
I think the second issue is that there are a lot of people, especially in the business
00:37:52.670 --> 00:37:58.050
world, but I think also, at least in certain countries in political and policy debates
00:37:58.050 --> 00:38:04.890
as well, who think that the risk of a crisis involving China, whether Taiwan focused, Philippines
00:38:04.890 --> 00:38:08.490
focused, India focused, is approximately zero.
00:38:08.490 --> 00:38:13.810
In their mental models, they assign a 0% probability and then they don't have to think about it.
00:38:13.810 --> 00:38:17.490
And I don't know what the probability is, but it's not zero.
00:38:17.490 --> 00:38:22.650
And so once you start thinking about the probability being 5% or 10%, suddenly the magnitude is
00:38:22.650 --> 00:38:28.670
so large, you've got to start incorporating the expected value into all of your calculations.
00:38:28.670 --> 00:38:33.790
And that I think is a second really important conclusion from participating in these war
00:38:33.790 --> 00:38:34.790
games.
00:38:34.790 --> 00:38:40.190
It doesn't take a sort of bolt from the blue or a real political surprise to reach a crisis.
00:38:40.190 --> 00:38:44.510
You can just follow forward current political dynamics between China and the Philippines,
00:38:44.510 --> 00:38:50.230
where there are Chinese maritime militia naval ships daily facing off against Filipino ships
00:38:50.230 --> 00:38:55.210
over disputed underwater island formations, shooting each other with water cannons.
00:38:55.210 --> 00:38:59.850
It's not hard to imagine how that could spiral in a very dangerous direction very quickly.
00:38:59.850 --> 00:39:04.850
I take your point a step further actually and say they assign a zero possibility because
00:39:04.850 --> 00:39:09.610
they don't want to think about what the potential outcome of a non-zero possibility would be
00:39:09.610 --> 00:39:11.530
of that happening because they don't want the answer.
00:39:11.530 --> 00:39:15.770
This is also why not to participate in the war game or thinking it through because that
00:39:15.770 --> 00:39:18.490
would require a level of change that they're not willing to deliver.
00:39:18.490 --> 00:39:22.430
And I think in Germany that goes for the chancellery and there's a serious question in Germany,
00:39:22.430 --> 00:39:27.390
are we seeing a divergence between those multinational companies, notionally German still, who are
00:39:27.390 --> 00:39:33.270
banking on being too big and too German to be allowed to fail, should the balloon go
00:39:33.270 --> 00:39:36.830
up in the Taiwan Strait, and Germany's people.
00:39:36.830 --> 00:39:39.390
Is that a dynamic that you see in other places as well?
00:39:39.390 --> 00:39:41.550
I mean, obviously this is a contested field.
00:39:41.550 --> 00:39:45.190
No, I think that is present in basically all of the G7 economies.
00:39:45.190 --> 00:39:49.850
You've got a smaller number of companies in basically every country that are making up
00:39:49.850 --> 00:39:53.450
a larger share of foreign direct investment into China.
00:39:53.450 --> 00:39:59.610
In Europe, the numbers are really quite shocking how half of FDI is produced by just 10 or
00:39:59.610 --> 00:40:01.930
so companies.
00:40:01.930 --> 00:40:06.690
And those companies, when they invest in China, the aim is to have future corporate profits
00:40:06.690 --> 00:40:09.050
coming from China.
00:40:09.050 --> 00:40:14.210
Those corporate profits accrue to the shareholders who are often international rather than German
00:40:14.210 --> 00:40:18.550
in the case of German companies or French in the case of French companies.
00:40:18.550 --> 00:40:22.970
And it does bring substantial risk that their investments could go to zero, just like the
00:40:22.970 --> 00:40:24.850
Russia investments did.
00:40:24.850 --> 00:40:29.730
Now, again, if you apply your 0% probability, you end up with some value on those investments.
00:40:29.730 --> 00:40:34.510
But if you apply a 10% probability or 20% probability, your scenario planning would
00:40:34.510 --> 00:40:36.410
look quite different.
00:40:36.410 --> 00:40:40.850
And that's why I think in addition to both governments and the policy community more
00:40:40.850 --> 00:40:45.370
generally putting a bit of pressure on companies to take risks seriously, to actually properly
00:40:45.370 --> 00:40:51.770
price risks, I think there is also a role for regulation and legislation in inducing
00:40:51.770 --> 00:40:59.370
companies to undertake similar types of risk assessments and scenario planning.
00:40:59.370 --> 00:41:04.650
And I do worry that there are some countries and some industries where CEOs have gotten
00:41:04.650 --> 00:41:09.330
away for too long with pricing risks at zeros and they're by taking, I think, irresponsible
00:41:09.330 --> 00:41:14.090
steps that, as you suggest, do eventually leave taxpayers potentially on the hook for
00:41:14.090 --> 00:41:15.090
bailing them out.
00:41:15.090 --> 00:41:19.290
Indeed, the pursuit of that private profit ends up with an externalized social risk for
00:41:19.290 --> 00:41:23.170
which we are exposed to, which are those who could deal with it and are not doing.
00:41:23.170 --> 00:41:26.990
And what we know from Germany is the China strategy in particular leaves all this very
00:41:26.990 --> 00:41:29.450
deliberately up to companies.
00:41:29.450 --> 00:41:33.890
And a lot of those companies are saying, thanks very much, zero possibility, off we go, as
00:41:33.890 --> 00:41:35.570
we were before.
00:41:35.570 --> 00:41:38.290
But there is investment in other ways as well.
00:41:38.290 --> 00:41:42.010
We see investment coming, for example, from TSMC into Germany.
00:41:42.010 --> 00:41:45.490
And we also see chip manufacturing in Europe being given a boost.
00:41:45.490 --> 00:41:48.770
Is your feeling that we're on the right track there or are we still in the business of picking
00:41:48.770 --> 00:41:50.970
winners in a rather unstrategic way?
00:41:50.970 --> 00:41:55.890
I think I do see the evidence of increased cross-border chip industry investments as a
00:41:55.890 --> 00:41:57.090
very positive sign.
00:41:57.090 --> 00:42:01.570
I think first it's a reminder that although it's true there's a number of Western businesses
00:42:01.570 --> 00:42:06.090
that invest a lot in China, Western countries invest a lot more in each other.
00:42:06.090 --> 00:42:08.690
And there's really no comparison.
00:42:08.690 --> 00:42:13.970
The amount of European FDI into the US or US FDI into Europe is substantially above
00:42:13.970 --> 00:42:16.970
either region's FDI into China.
00:42:16.970 --> 00:42:19.890
That's critically important to keep in mind.
00:42:19.890 --> 00:42:24.170
I also think that I occasionally hear people talk about de-globalization.
00:42:24.170 --> 00:42:26.210
And in the chip industry, I see no evidence of that.
00:42:26.210 --> 00:42:31.090
I see US firms investing in Ireland and Germany in Israel.
00:42:31.090 --> 00:42:34.210
I see Japanese firms investing more in Taiwan.
00:42:34.210 --> 00:42:38.170
I see Taiwanese firms investing in Japan and the US and Germany.
00:42:38.170 --> 00:42:42.330
And so de-globalization does not remotely describe what we're seeing right now.
00:42:42.330 --> 00:42:48.490
I think we are seeing a deliberate effort to reduce reliance on China, that's for sure.
00:42:48.490 --> 00:42:53.330
But I think rightly so, both businesses and governments have tried to make this not about
00:42:53.330 --> 00:42:55.490
self-sufficiency, not about de-globalization.
00:42:55.490 --> 00:43:02.130
And we're beginning to see in at least certain segments creation of a political block that
00:43:02.130 --> 00:43:03.610
maps onto an economic block.
00:43:03.610 --> 00:43:08.050
And to me, that's a world that seems much less risky than the one we currently live in.
00:43:08.050 --> 00:43:09.050
Right, absolutely.
00:43:09.050 --> 00:43:12.890
It's that rethinking of what do we mean by the globe in globalization and who's part
00:43:12.890 --> 00:43:16.850
of that or not that I think is one of the key geopolitical and geo-economic acts of
00:43:16.850 --> 00:43:18.010
this time.
00:43:18.010 --> 00:43:21.930
One thing that struck me in the book was not only was I actually able to now finally understand
00:43:21.930 --> 00:43:25.290
what a silicon chip is, what a semiconductor is composed of.
00:43:25.290 --> 00:43:27.650
So you managed to make the technology understandable.
00:43:27.650 --> 00:43:29.130
But also you managed to humanize it.
00:43:29.130 --> 00:43:32.810
And you did so through talking through the stories of the players involved.
00:43:32.810 --> 00:43:38.410
And there really were some characters involved in this in the business of creating semiconductors
00:43:38.410 --> 00:43:42.090
and inventing them in the first place, as well as then in marketing them.
00:43:42.090 --> 00:43:45.930
Could you talk us through some of your favorite characters you came across in this book?
00:43:45.930 --> 00:43:51.570
And I'm thinking of people like Bob Norris, Andy Grove, Pat Haggerty, Morris Chang.
00:43:51.570 --> 00:43:57.770
Morris Chang is the person who is most underestimated in terms of his importance and who illustrates
00:43:57.770 --> 00:44:01.010
in a lot of ways the development of the chip industries.
00:44:01.010 --> 00:44:06.610
One in mainland China before the revolution fled when the communists took power in '49
00:44:06.610 --> 00:44:11.490
enrolled at Harvard, then MIT, and was really present at the creation of the chip industry.
00:44:11.490 --> 00:44:16.290
Worked for Texas Instruments for many decades, held a top secret US security clearance for
00:44:16.290 --> 00:44:22.810
his work on defense electronics, and then moved to Taiwan in 1987 to found a TSMC and
00:44:22.810 --> 00:44:25.770
helped it grow to become the world's largest chip maker.
00:44:25.770 --> 00:44:30.450
And so I think his life illustrates on the one hand the really deep connections between
00:44:30.450 --> 00:44:35.090
the chip industry and East Asia that were present from the beginning, represents the
00:44:35.090 --> 00:44:43.050
globalized nature of the chip industry, and also represents, I think, some of the challenges
00:44:43.050 --> 00:44:49.650
that the West faces in trying to navigate its reliance on TSMC and Taiwan.
00:44:49.650 --> 00:44:54.130
Because there are people like him who are not replaceable.
00:44:54.130 --> 00:44:59.050
They have human capital and expertise and networks that make it very difficult to imagine
00:44:59.050 --> 00:45:04.650
shifting the geographic footprint of the chip industry in a short period of time.
00:45:04.650 --> 00:45:06.490
And that's really critical to remember.
00:45:06.490 --> 00:45:11.370
Chips are built by machines and by factories, but they're ultimately built by people.
00:45:11.370 --> 00:45:17.050
And where those people live and where their networks exist is just as important as any
00:45:17.050 --> 00:45:19.850
decision about investment policy in terms of shaping the industry.
00:45:19.850 --> 00:45:20.850
Right.
00:45:20.850 --> 00:45:24.330
And indeed also where those people got their habitats, where they got their founding qualities
00:45:24.330 --> 00:45:28.210
of their own lives, their situated folk in particular places, of course.
00:45:28.210 --> 00:45:34.170
And Chang's background as an immigrant into the US and then a migrant into Taiwan again,
00:45:34.170 --> 00:45:38.250
a bit of a Steve Jobs figure in some ways that he retired from the business, then had
00:45:38.250 --> 00:45:41.490
to come back to save it, drag it back up to his value base.
00:45:41.490 --> 00:45:46.730
But there was also another migrant, Andy Grove, Andras Groff, who had a fascinating story.
00:45:46.730 --> 00:45:48.730
Perhaps you tell us a little bit about him.
00:45:48.730 --> 00:45:55.930
He was born in Budapest, fled after the communists took power as well, arrived penniless in New
00:45:55.930 --> 00:46:04.970
York in the mid fifties, got a degree in engineering and started working at Intel right when it
00:46:04.970 --> 00:46:11.010
was being founded in the late 1960s and ended up becoming the CEO of Intel throughout the
00:46:11.010 --> 00:46:17.050
1980s and made it at the time the world's largest and most successful chip firm.
00:46:17.050 --> 00:46:23.410
And I think he illustrates alongside Morris Chang the extent to which the chip industry
00:46:23.410 --> 00:46:27.330
really has always drawn from an international talent pool.
00:46:27.330 --> 00:46:33.650
If you look at the founders of key chip companies in the US, for example, or in Asia, if you
00:46:33.650 --> 00:46:39.130
look at the key innovators behind technological improvements, new transistor designs, new
00:46:39.130 --> 00:46:43.370
chemicals, a really surprising share are immigrants.
00:46:43.370 --> 00:46:47.970
And I think that's a dynamic that is very important to keep in mind today as governments
00:46:47.970 --> 00:46:51.570
tighten up rules over talent flows, for example.
00:46:51.570 --> 00:46:56.530
They need to make sure they don't tighten them too far and limit the ability to attract
00:46:56.530 --> 00:47:00.850
talent that still remains, I think, key to retaining on top psychologically.
00:47:00.850 --> 00:47:01.850
That's it.
00:47:01.850 --> 00:47:04.450
We're in a competition for talent as well as a geopolitical competition and the two
00:47:04.450 --> 00:47:05.530
are highly connected.
00:47:05.530 --> 00:47:09.770
So to leverage the advantage of our free societies, we actually have to let freedom breathe.
00:47:09.770 --> 00:47:10.770
But tell me one more thing.
00:47:10.770 --> 00:47:12.130
Where are the women in this story?
00:47:12.130 --> 00:47:17.090
Well, the chip industry deserves a fair amount of criticism for the fact that it has been
00:47:17.090 --> 00:47:18.890
extraordinarily male dominated.
00:47:18.890 --> 00:47:23.850
And actually, women have played a very large role in chip production processes, but largely
00:47:23.850 --> 00:47:26.650
in the assembly and the test proportions.
00:47:26.650 --> 00:47:33.850
If you look at the factories that were assembling chips in the early decades of the chip industry,
00:47:33.850 --> 00:47:36.810
it was largely women on the assembly lines.
00:47:36.810 --> 00:47:40.890
But it was not largely women in the design process or in the leadership positions.
00:47:40.890 --> 00:47:45.490
And even today, I think the chip industry is, I would give it low marks compared to most
00:47:45.490 --> 00:47:51.210
other industries in terms of women in the executive positions or prominent female leaders.
00:47:51.210 --> 00:47:56.690
I think Lisa Sue of AMD, current CEO, is one very important exception.
00:47:56.690 --> 00:48:01.050
But I think the chip industry has a long way to go when it comes to achieving gender equality
00:48:01.050 --> 00:48:02.050
goals.
00:48:02.050 --> 00:48:06.330
Again, another key element of leveraging our advantage, just to round that out, what I
00:48:06.330 --> 00:48:10.330
found particularly admirable or one of the many things I found admirable about Chris Miller's
00:48:10.330 --> 00:48:14.610
book, Chip Warp, is the way that he follows and traces the story of those characters as
00:48:14.610 --> 00:48:20.010
a way of tracing technological development, geopolitical competition, the economic ups
00:48:20.010 --> 00:48:24.450
and downs of particular firms, but also countries involved.
00:48:24.450 --> 00:48:29.450
And this process of rebuilding Intel that Andy Grove undertook, which detailed in his
00:48:29.450 --> 00:48:33.170
famous book called Only the Paranoid Will Survive, great book title.
00:48:33.170 --> 00:48:35.530
This is a recurring cycle, of course, in business.
00:48:35.530 --> 00:48:39.890
And we have to be able to take the stories from that, take the lessons from that forward.
00:48:39.890 --> 00:48:43.090
But keeping sight of the people involved, I think, is something that Chris does particularly
00:48:43.090 --> 00:48:48.810
well for someone who's operating at that level of geopolitical and geoeconomic sophistication.
00:48:48.810 --> 00:48:52.210
One small example close to the beginning of the book, we introduced to a person called
00:48:52.210 --> 00:48:57.770
Jay Lathrop, who actually invented the lithography techniques that are now absolutely crucial
00:48:57.770 --> 00:49:03.930
to producing silicon chips, to producing the semiconductors that are in all of our technology.
00:49:03.930 --> 00:49:07.050
It was recognized at the time, this was such a breakthrough, that he was still actually
00:49:07.050 --> 00:49:10.170
working for the army before he went to go and take a corporate job.
00:49:10.170 --> 00:49:14.890
And the army paid him $25,000 in cash for this discovery.
00:49:14.890 --> 00:49:17.090
And that was back in the late '50s.
00:49:17.090 --> 00:49:21.170
And this is a technological breakthrough that changed the world as we know it.
00:49:21.170 --> 00:49:26.850
At the very end of the book, Chris comes back into this story of Jay Lathrop, talking about
00:49:26.850 --> 00:49:32.210
moving his family to Texas, to Austin, to work for Texas Instruments.
00:49:32.210 --> 00:49:34.290
And we're told they were still in the old car.
00:49:34.290 --> 00:49:38.250
They hadn't actually brought the new station wagon, which that $25,000 would eventually
00:49:38.250 --> 00:49:40.090
give to the family.
00:49:40.090 --> 00:49:43.930
So it's picking up all the way through those developments that are personal, as well as
00:49:43.930 --> 00:49:47.690
economic, political, and technological, and scientific.
00:49:47.690 --> 00:49:51.570
So highly recommend to read Chris Miller's book, Chip War, and we'll be following up
00:49:51.570 --> 00:49:56.170
on some of those themes in the remainder of this season of Burlinside Out and in our future
00:49:56.170 --> 00:49:57.170
endeavors.
00:49:57.170 --> 00:49:58.170
Absolutely.
00:49:58.170 --> 00:50:01.010
Thanks so much to Chris Miller, author of Chip War, for joining us.
00:50:01.010 --> 00:50:06.890
And one thing that is very clear in both the book and also our chat with him today is that's
00:50:06.890 --> 00:50:07.970
why he calls it Chip War.
00:50:07.970 --> 00:50:12.450
The systemic competition element that is at work here.
00:50:12.450 --> 00:50:17.410
And we heard a lot in this particular episode about how that's relevant when we consider
00:50:17.410 --> 00:50:25.930
China and the threat that China poses to Taiwan in particular, but also to the security of
00:50:25.930 --> 00:50:30.290
really the globe and our own Western societies in its own various ways.
00:50:30.290 --> 00:50:32.290
But there's another element to this as well.
00:50:32.290 --> 00:50:34.810
It's not simply the threat that China poses.
00:50:34.810 --> 00:50:39.370
There's a little bit of an indirect thing that we can follow here when it comes to the
00:50:39.370 --> 00:50:41.650
systemic competition elements of chips.
00:50:41.650 --> 00:50:46.450
And that is that there is actually a Russian element to this as well.
00:50:46.450 --> 00:50:51.810
The Russian military is getting access to Western chips usually actually by buying them
00:50:51.810 --> 00:50:53.450
from China.
00:50:53.450 --> 00:50:59.850
And those chips are obviously being used to wage war against Ukraine.
00:50:59.850 --> 00:51:04.850
And that is obviously has huge implications for our own security.
00:51:04.850 --> 00:51:08.970
So when we think about this chip war, there's other theaters that it pops up, not simply
00:51:08.970 --> 00:51:11.610
in East Asia, of course.
00:51:11.610 --> 00:51:19.850
So when we're talking about going for victory in a place like Ukraine, another thing that
00:51:19.850 --> 00:51:25.170
we really need to consider, I would say, is the tightening of sanctions and export controls
00:51:25.170 --> 00:51:26.330
on these chips.
00:51:26.330 --> 00:51:31.730
How are these Western chips that are in China going to Russia and directly affecting the
00:51:31.730 --> 00:51:33.690
war in Ukraine and our own European security?
00:51:33.690 --> 00:51:38.890
We have to be conscious of this in all aspects because we've seen the use that chips can
00:51:38.890 --> 00:51:43.890
be put to be that re-equipping dumb bombs to make them into smart bombs, guided bombs,
00:51:43.890 --> 00:51:47.770
which the Russians have been using, even taking things, chips from washing machines or so
00:51:47.770 --> 00:51:48.770
on.
00:51:48.770 --> 00:51:52.610
So we really have to have a global view of this to make sure that we are restricting
00:51:52.610 --> 00:51:58.930
Russia's capability to wage war and certainly stopping its capability to win war on all
00:51:58.930 --> 00:52:00.770
the fronts where it exists.
00:52:00.770 --> 00:52:04.490
That's why we have to consider chip war in a full global context.
00:52:04.490 --> 00:52:06.410
That's all for this episode of Berlin's Side Out.
00:52:06.410 --> 00:52:09.010
Thank you very much to our guest today, Chris Miller, for joining us.
00:52:09.010 --> 00:52:13.330
You can find information on his book, Chip War, in our show notes.
00:52:13.330 --> 00:52:17.330
Thank you also to our project assistant, Yulia Stukla, our technical producer, Hendrik Banna,
00:52:17.330 --> 00:52:19.210
and the DJP team.
00:52:19.210 --> 00:52:23.810
Joining us over the next few weeks, we had to talent for the Leonard Mary Conference
00:52:23.810 --> 00:52:27.530
and bring you the talk of the town straight from there, as well as our upcoming episodes
00:52:27.530 --> 00:52:32.530
on what's needed for Ukrainian victory, the pillars of victory, if you will, and the
00:52:32.530 --> 00:52:37.650
role Russia's nuclear blackmail is playing into current German and Western strategic
00:52:37.650 --> 00:52:39.050
choices.
00:52:39.050 --> 00:52:42.410
For now, though, from Berlin, Auf Wiedersehen and Tschüss.
00:52:42.570 --> 00:52:45.610
[MUSIC PLAYING]
00:52:45.610 --> 00:52:48.970
[MUSIC PLAYING]
00:52:48.970 --> 00:52:52.330
[MUSIC PLAYING]
00:52:52.330 --> 00:52:55.690
[MUSIC PLAYING]
00:52:55.690 --> 00:53:05.690
[BLANK_AUDIO]