Sputnik, AI, and the Nature of Victory

The US foreign-policy community has been gathering itself around the goal of winning the AI race against China. The problem is that defining “winning” is not at all easy. If winning consists of US companies, in cooperation with the US government, enjoying a monopoly on the best AI technology for some extended period — which does seem to be what is expected — SIG’s view is that winning is nearly impossible. The only way the US could come close is by sharing technology within some type of alliance. But that would entail non-American companies within the alliance having revenues and profits of their own. The US and US companies cannot “win” this alone.

As SIGnal has emphasized before, digital technology has been taking the world’s defense sectors by surprise for some 30 years. Whether it is low-earth-orbit satellite swarms, drones or navigational improvements, technology developed for one use becomes a military must-have for security uses. Proliferation is built into such a process. Military hardware needs software; software lends itself to proliferation, theft, imitation, and improvement. Artificial-intelligence software is no different.

Containment of American AI within US boundaries goes against the nature of the 21st-century technology industry. Most innovation comes from the private sector, whose ability to maximize profit and minimize costs depends on a global marketplace for products and labor. The defense sector is not the private sector but a curious public-private blend. American defense companies do sell a lot to overseas customers, but the customer whose needs shape the greater part of production is the US government. Proliferation of American defense contractors’ products, including software and data, is carefully regulated. Workers need to get government clearances. Contracts have to conform to official bureaucratic standards. There is plenty of red tape. The payoff for defense companies has been the security of long-term contracts and a relatively high level of protection from competition — notably from foreign competition.  The main downside is that profits from such quasi-public business, in the absence of corruption and favoritism, are limited by the obligation of Congress to ensure that government is not over-spending. Innovation within the defense sector thus seems to come up against natural limits. That is not the case in the private sector, which is why so much military innovation comes from outside the defense sector and commonly occurs for reasons that have nothing to do with defense.

This is abundantly true of AI innovation. If the US government wanted to make AI innovation henceforth a government-controlled process, it would amount to turning AI companies into defense companies — which would remove much of their incentive for innovation, defeating the purpose of the exercise. It would not be much of a victory in the race for AI dominance.

By contrast, operating with trusted partner countries would have some of the advantages of globalization — multiple labor and consumer markets to choose from — while preserving the goal of excluding China and other antagonists. Of course, forming some sort of digital alliance structure has been a US goal since the middle of the first Trump administration. Results have been mixed. There has been a contradiction at their core: The US wants partners but insists on being the dominant one. That kind of dominance cannot work in the case of private-sector-led technology innovation.

Fortunately US tech companies, although in their own ways just as hungry for dominance as the US government, have become accustomed in the last decade to competing in markets with foreign companies and not always winning. They have invested huge amounts in overseas markets: to pay suppliers, establish their own production, or attract customers but also to take advantage of the huge and growing innovation ecology that exists outside the United States. And foreign governments and private competitors have gotten used to them as well. The degree to which US tech companies can be profitably active in non-American markets without dominating them is an example of a type of loose alliance. The struggle with China is an important shaping factor but it does not distort everything it touches.

Learning from the success of this private-sector-led approach to the US-China tech contest could lead to a public-sector variant that could help control AI proliferation while accepting that winning the AI race with China, in the winner-take-all sense, cannot be done. A different type of victory might be possible though. After all, when the US, following the Soviets’ shocking Sputnik launch in 1957, went all out to win “the space race” against the USSR, it did not so much prevail as demonstrate its ability to continue to innovate at a pace the Soviet Union could not match. The result, in 1975, was American and Soviet astronauts living together in the International Space Station (as Russians and Americans still do) and the growth of an international scientific subculture that played an important role in bringing the Soviet experiment in oppressive governance to a close.   

AI is Just a Tool

By Dee Smith

There are many problems with AI, some of which I will explore in future posts. But the most basic problem is that, as we have all experienced, computers break.

For computers to continue to run requires multiple people who are capable of fixing them, available all the time.

Remembering this, is it a good idea to give more aspects of our lives over to “intelligent” systems so undependable? The things we rely on to obtain the food we eat, the water we drink, and to make, manage, and spend our money? The systems we use to conduct business, to take care of our health, our critical infrastructure, and our national security?

We already do, of course, but the teams are in place to fix them when they malfunction.

The unreliability of computers is not a passing problem. Computer systems, considered as a whole, are scarcely more reliable now than they were 30 years ago. Hardware is somewhat more reliable, but software is increasingly complex, increasingly unpredictable (complex systems are inherently more unpredictable), and increasingly unreliable.

Relying on AI systems makes us vulnerable in several critical ways. First is their exposure to attack. To cite just one example: discovery of undetected flaws leading to “zero-day exploits” — criminal or terrorist attacks exploiting those flaws.

Second are the continuing “hallucinations” AI experiences, where it gives entirely wrong, and sometimes nonsensical, information, often for reasons computer scientists do not understand. What if it does this while managing an element of critical infrastructure and the problem is “inside” the system, where it cannot easily be detected or fixed?

Third, all computer systems are subject to severe malfunctions due to rare, but potentially catastrophic, single-event upsets (SEUs) or single-event errors (SEEs) caused by cosmic rays bombarding the earth.

Fourth is AI’s requirement for a vast and ever-increasing level of electrical power for operation.

The reason computer systems are so ubiquitous is, of course, money. This works in two ways: the money being made and the money being saved by replacing human laborers. From a social standpoint, the latter may well be a pyrrhic victory: displacing millions of people from their jobs creates a huge social cost, in real money.

Are computer systems, in general, more efficient than humans? There is no evidence that they are. Computers are able to crunch numbers within mathematical operations much faster than humans — although that is discounting the enormous calculational power of the brain of a human, let alone the brain of a bird or even an ant, doing everyday things. There is no real understanding of how these biological intelligent systems work. Computer systems seem more efficient only because of the extremely limited scope within which they are operating.

Consider two alternatives, at opposite ends of the spectrum. One is that computer systems, as they become more and more complex, also become more and more fragile. When a system related to food production, or finance, or national security breaks catastrophically somewhere, the failure cascades through the system.

What if systems could be made substantially more reliable? Perhaps some unforeseen breakthrough will dramatically improve their dependability. Then suppose, as some people insist (incorrectly to my thinking), that AI can and will progress to Artificial General Intelligence (AGI). Imagine that this results in a superhuman intelligence. It could be one that emerges at a critical-mass-type point, almost in an instant (this is called the “singularity” by AGI aficionados). Were this to happen, we have no way of knowing whether such an entity would be benign, neutral, or malicious to humans.

But if such an AGI is trained on the sum total of human knowledge and expression, then that AGI is going to be loaded with all the bad along with the good. Do we really want to live in a world governed by transcendently intelligent and powerful machines trained on the behavior of what are essentially clever, volatile, often enraged chimpanzees? (We share 98.4 percent of our DNA with chimps.) Watching any war movie, or really most any movie, would suggest we might not.

And if the AGI was not trained on human knowledge and culture, what would it be trained on?

Biological systems have had about 4 billion years of evolution on this planet to become reliably dependable in operation. They are generally able, as living systems, to survive constant bombardment by radiation from space, extreme temperatures, rapid changes in climate, changes in atmospheric chemistry — and most important, to survive without someone standing by to repair or reboot them. This is a property known as homeostasis. Life has evolved naturally over an immense period of time through adaptation: trial and error.

One the other hand, our computer systems — based on silicon, not carbon — do have a very fallible creator: us. And they have been around about 70 years, or about two-trillionths as long as biological systems.

The belief in the inevitable ascendence of AGI is an article of faith for many involved in the computer industry and for others outside the industry who uncritically accept this “techno-religious” belief system. In its more virulent forms, it is teleological: a burning faith in an inevitable direction of history, in which AGIs are the successors to humanity. And in which the sacred duty of computer scientists is to bring about the birth of this supremely intelligent “life” form.

If I had told you 30 years ago that you would have in your pocket a self-powered device the size of a pack of cards that could tell you how to drive, turn by turn, from your current address to a building in a city 1000 miles away, you would probably have thought that it must be intelligent to be able to do this.

Do you think of your smart phone that way today? My estimation is that this is how we will think of AI in 30 years: a useful, not entirely dependable tool. Nothing more.

The Rest Is Software

US President Donald Trump’s visit to the Persian Gulf brought the region back into the American camp on artificial intelligence. The White House’s cancellation of the Biden administration’s AI-diffusion regulation was well timed: the message of both the trip and the cancellation was that this administration will not draw distinctions, as its predecessor did, in advancing what Commerce Secretary Howard Lutnick called “Trump’s vision for US AI dominance.” The US is, in a sense, trying to de-regulate AI politically. Washington’s move to block AI regulation by US states is also part of this. In SIG’s view, whether such de-regulation will achieve the goal of AI dominance is a different question.

As with crypto, the current administration’s US’s bias with AI is to let the chips fall where they may, so to speak, while also aggressively using the power of the state — as investor, as enforcer, as customer — to secure American advantages. Trump’s experiences of being deplatformed by Big Tech must have shaped his views: bitterness over the suppression of conservative speech, alongside the supposed promotion of anti-conservative speech, has been a dominant note since his second inauguration. In this scenario, technology and tech innovation were shown not to be autonomous forces, proceeding according to their own logic, perhaps capable of being channeled but not of being controlled. Rather they were the effects of companies run by individuals who could be influenced. That was well within the comfort zone of a lifelong businessman. (See the tariff retaliation against Apple for relocating its China production in India rather than the US.) It is a pro-market perspective in a way, but with the market understood as a place for ruthless competition among a small number of unconstrained players rather than as a mechanism for maximizing the efficient distribution of capital and labor.

Similarly, the role of the state in this perspective is to personify the nation in unconstrained and ruthless competition among states for, in U.S. Commerce Secretary Lutnick’s term, “dominance.” President Trump’s appetite for military confrontation in his first term was low, and that seems to be carrying into his second term. His appetite for economic confrontation was relatively high in term one and has gone to a new level in term two. The tools of the state are the weapons he has for such confrontation. They are directed toward securing dominance. Trump is personifying the powerful idea of economic nationalism.

The difficulty, with regard to “US AI dominance,” is that the AI sector is not like other industrial or commercial sectors. The preferred means for dominating AI has been the control of hardware, as in export controls on leading-edge chips or chip-design lithography equipment. Biden’s AI-diffusion regulations, like his CHIPS Act and much else, were about the geopolitics of hardware distribution. President Trump has opened that floodgate. But once the hardware starts flowing and the data centers are built the rest is software, the diffusion of which is extremely hard to control. Software can be stolen or replicated; more important, it can be developed independently, as DeepSeek has shown. The supply of chips and what is necessary to manufacture them can be choked off, up to a point. The supply of engineers and software-engineering skills really cannot. It will be diffused regardless of what the US or China want.

Among other things, this means US AI dominance depends on the strength and autonomy of US universities, the freedom to innovate in the US tech sector independent of political agendas, the smooth functioning of open global markets, sensible market pricing of resource inputs, the reduction of obstacles to the cross-border movement of labor … all of which run contrary to current US policy.

The Gulf states are investing in US AI infrastructure on the way to building their own systems, which will have the capacity to become independent of US systems (see SIGnal, “The America Stack,” Feb. 5, 2025). The emiratis are not happily volunteering to be hostages to US AI dominance. They are seizing the opportunity to gain access to the best technology that will enable them to maximize their own sovereignty while positioning themselves to be a sort of port for the storage, manipulation, and distribution of data, just as Dubai’s port operates with coffee, tea, and so much else.

The pattern is similar elsewhere, although no one can direct capital with quite the speed, and in quite the volume, that the Gulf states bring to bear. Malaysia hesitated for a moment at new deals for Chinese technology when Washington threatened retaliation against states using Huawei’s latest AI chips, but in the end, the shape of AI is not going to be determined by hardware. The massive computing power required to participate in the search for the grail of Artificial General Intelligence (AGI) is indeed a hardware question, but for sub-AGI artificial intelligence, which might well prove to be most if not all of AI, hardware is only one factor. The rest is software. And US dominance of it is unlikely to be secured using the current means.

Can AI Make a Country Great Again?

Much recent commentary on artificial intelligence (AI) has focused on the prospect of a company or a country winning a race for artificial general intelligence (AGI) or more-than-human “superintelligence.” However, that goal, which seems rather more religious than technological, is both elusive and, should it ever be achieved, fragile (see SIGnal, “Mutual Assured Malfunction,” March 13, 2025). Investors are focusing instead on “little tech” and firm-level or industry-level AI that uses specific data sets to engineer specific productivity gains. In SIG’s view, this more modest course seems both economically more promising and politically much more sustainable. But it definitely does have risks of its own.

The appeal of “little tech” AI is partly that it leaves to one side the many serious questions about data privacy and other more existential matters that are posed by AGI. Smaller AI systems can run on the contained, often proprietary data sets involved in industrial processes, especially in manufacturing. The goal is not to replicate the human brain but to make industrial processes more efficient, raising productivity. It is a type of automation, using new technology yet still familiar enough from the history of industrial production.

With little-tech AI, startups can focus on specific problems whose solutions will provide a payoff in the relatively short term. In other words, AI would be monetizable. This has an obvious appeal not just to startup investors but also to industrial incumbents whose processes would be improved and whose productivity would be raised in competition with their rivals. Startups are not alone in this sphere. The German giant Siemens, for example, has put industrial AI at the core of its offering.

Politically, this approach to AI is much more appealing to most governments, only a few of which (the US, China) can have much hope of achieving global dominance by winning a race for AGI, at which point they might well regret getting what they wished for. Leaving aside the large question of AI data-center electricity demands, it offers the attractive prospect of raising productivity while reducing carbon use — because your factory in Texas, enhanced by AI, will no longer have to source so many of its components from East Asia, with all the carbon-using transport that entails. The little-AI approach also means states would not have to expose their citizens’ data to foreign tech multinationals, possibly based in hostile or overweening states, in order to participate in the later 21st century. That would be a gain for state sovereignty; and given that so many of the tensions around globalization have had to do with the way it threatens sovereignty and the democratic (or otherwise) accountability of governments to citizens, the little-AI approach could conceivably enhance global stability and the prospects for peace. Little AI, by improving productivity within a given national domestic workforce, could help states that are facing demographic stagnation — which is pretty much all industrialized states and many less-industrialized ones — to nonetheless grow on the basis of domestic labor (see SIGnal, “AI Family Values,” May 3, 2024). As Marc Andreessen and Ben Horowitz wrote in July 2024, “little tech” could make it possible “to reconstruct the American manufacturing sector around automation and AI, reshoring entire industries and creating millions of new middle class jobs” while also having green benefits. Technology could, in effect, provide the “labor” that would solve the biggest challenge facing President Trump’s vision of a more self-sufficient US: the lack of workers operating at a sufficient level of productivity (see SIGnal, “Trade Wars and US Labor,” April 11 2025).

Less carbon use, stabilization of the international sovereign-state system, a growing middle class, a renewal of rich-world domestic manufacturing but with higher wages and less grim manual work…What could possibly go wrong?

AI-enhanced production aimed at reshoring manufacturing to high-wage economies would square the circle of productivity growth and de-globalization. It would revive the pre-1975 global industrial status quo with the crucial addition of China (but not so much India or Southeast Asia). If you have the good fortune to live and work in a benefitting state, this would be a positive outcome. It could, however, also fuel techno-nationalism in the rich world (plus China) and make growth outside the AI-enhanced nations highly problematic. One key issue raised by the US-China struggle — a protected US market deprives non-American producers of consumers, while a protected Chinese economy, likewise deprived, dumps its production for the pre-tariffs US market onto the rest of the world’s economies — would be gravely worsened as the world’s two largest economies reduce their dependence on the rest of the world for both supply and demand.

AI-enhanced de-globalization could, in short, reverse the global redistribution of labor productivity that led to the greatest poverty reduction in human history. In theory, the gains from little AI could be more equally distributed. After all, the AI enhancements that would lift an underemployed person in Oklahoma or eastern Germany into the middle class of his or her domestic economy could do the same for a person in Nigeria or Thailand. But that outcome is not the goal for the people, states, and companies that are driving the growth in AI monetization. Their goal is nearly the opposite. For investors, the greatest gains will come from identifying companies and sectors best positioned to gain from AI-enhanced de-globalization.

Mutual Assured Malfunction

The past week has been a lively one for the eternal battle between digital networks and national, sovereign security. After a two-year standoff, Elon Musk’s Starlink was able to reach deals with India’s #1 and #2 telecommunications companies, Reliance Jio and Bharti Airtel, on providing satellite Internet to the subcontinent’s vast and underserved rural market. A few days earlier, Dan Hendrycks, Eric Schmidt, and Alexander Wang — respectively, director of the nonprofit Center for AI Safety, former chairman of Google, and the CEO of Scale AI — released a paper , “Superintelligence Strategy,” arguing that no one state will ever be able to win the AI race.

In the first instance, a technology company with, it is fair to say, its own distinctive geopolitical interests could potentially gain a hold over the telecommunications of the world’s second largest national market. In the second instance, tech industry leaders with, particularly in the case of Schmidt, a strong record of advocating US technology dominance in competition with China are asserting that such dominance can never be complete. Indian digital sovereignty and US digital sovereignty are rendered highly problematic if not impossible. If a state is on the networks, as all powerful states are and will be, then their sovereignty is inherently partial. Taking these two major developments together, the future of digital self-determination can be seen to be rather weak. In SIG’s view, this represents an overdue recognition of the interdependence of states even as they engage in fierce geopolitical competition.

Reliance Jio has, in the past five years, revolutionized India’s telecommunications, particularly mobile communications, bringing huge numbers of Indians online. Bharti Airtel has done a surprisingly good job at catching up, giving Reliance Jio much-needed competition. The Indian state has not been idly observing these developments. Its vigorous advocacy of an indigenous digital infrastructure, often now referred to as the “India Stack,” has become an example to others, including the European Union. (See the SIGnal post “The America Stack,” Feb. 5, 2025.) India is determined to become a major tech power. It has also, with the world’s fourth-largest defense budget after the US, China, and Russia, aggressively advanced its own space program and its own space-based navigational system to rival GPS (US), Glonass (Russia), and BeiDou (China). Balancing US and Chinese telecommunications majors over the past decade-plus, India has artfully and purposefully pursued its desire to achieve digital self-determination.

That made the Starlink deal a surprise. It appears to have been hammered out between Musk and Indian President Narendra Modi during the latter’s recent visit to Washington. The Indian government has an interest in nurturing Reliance Jio and Bharti Airtel, but it also has an interest in good relations with the US under President Donald Trump and in making sure that neither Reliance nor Airtel accumulates too much power domestically. Both the US and China have faced a similar problematic in simultaneously backing and controlling their own tech majors. The deal with SpaceX, Starlink’s parent company, provides one way for India to meet these several challenges. Indian reaction to the Starlink deal has been understandably wary and somewhat confused. After all, the Indian government, at various junctures, has humbled Facebook, Google, Amazon, Microsoft, Huawei, and ZTE, among other foreign firms eager to reach the Indian market. A recent Indian report characterized Starlink as “a technology of geopolitical control,” pointing meaningfully to Starlink’s role as the guarantor and master of Ukraine’s Internet access in that country’s struggle with Russia.

SIG’s view is that Starlink will not be able to repeat its Ukraine dominance in India, any more than its US and Chinese rivals have been able to subdue the subcontinent — not for want of trying. It is nonetheless striking that Modi, Reliance, and Airtel — the latter two have long opposed letting Starlink into the tent — now believe that the advantages of working with SpaceX outweigh the disadvantages. At the very least, Musk has dramatically proved that having the ear of the US president provides enormous business benefits.

While the “Superintelligence Strategy” has been in the works for some time, it is difficult not to read it in the context of the Trump administration’s declared determination to press the US’s AI dominance. One of Trump’s first moves was a $500 billion AI infrastructure project, and Vice President J.D. Vance later stressed in a landmark speech in Paris that the US “will ensure that American AI technology continues to be the gold standard worldwide and we are the partner of choice for others — foreign countries and certainly businesses — as they expand their own use of AI.” Vance held back from a simple declaration of hegemony but the administration’s message has clearly been that US AI should indefinitely be the parent in comparison to the efforts of other nations, especially China.

The “Superintelligence Strategy” has at its core the highly convincing argument that any large-scale AI system, even an American one, will always be vulnerable to infiltration and disruption by rivals. The strategy offers a very worldly solution, based on, but distinct from, earlier strategic arguments about nuclear weaponry. It is called Mutual Assured AI Malfunction (MAIM): the acceptance that there will be a balance of AI power, not a resolution or well-meaning regulation of it. Further, MAIM “already describes the strategic picture AI superpowers find themselves in.” A new Mutual Assured Destruction (MAD), AI version, is already with us. As in the earlier, nuclear version, there can be no victors.

There is much here for China and others to digest. The old, US-led idea of a free and open Internet, so recently repudiated, can be seen as returning, but in a much darker form appropriate perhaps for darker times. How states and companies react is the crucial question for investors. The venerable commercial goal of scaling, ideally to a global level, is not going to be achievable. AI-fueled tech companies, which increasingly means most tech companies, will face geopolitical limits. Commercial cooperation within those limits — and successful digital competition is inherently commercial — seems to be the only way forward. Musk, Modi, and the authors of the “Superintelligence Strategy” are simply ahead of the curve, and showing the rest of us where it bends.

The America Stack

Investing in technology got a lot harder in the past two weeks. Tech investors, particularly in AI, have traditionally assumed the best products would scale: pick the winner, and you will win big. That assumption will sometimes still prove valid, but it now seems fundamentally outdated. Technology markets are fragmenting for reasons that are not changing soon. That is making investors’ lives difficult.

The proximate cause for the drop in US tech stocks was DeepSeek’s launch of AI products that seemed to perform tasks that the company’s American competitors do at much greater cost. A Chinese company whose actual workings are opaque even by Chinese standards, DeepSeek surprised markets. The specific instance was indeed unanticipated, but the broader phenomenon, as SIGnal readers know, should not have been. The US has been tightening the screws on Chinese technology for years. The first Trump administration took technology containment to a new level and the Biden administration went further still. Neither administration explained what a realistic endgame was. But it was obvious that China and Chinese companies were not simply going to yield and give up. Every US sanction and prohibition has been met with Chinese innovation. The resulting products might not match their US analogues byte-for-byte, but they don’t have to. They just have to be good enough to enter the markets. Then they can win on price.

DeepSeek’s ability to do that burst the AI bubble, which was inflated by confidence in the US tech sector’s ability — supported by government spending and other encouragements — to prevail on the global scale. That confidence is now weakening, not just because a Chinese tech company can compete with America’s best but because the “global scale” has been shown to be a fiction. Neither of the world’s two largest economies is going to either give up on protecting and subsidizing its tech companies or open its digital markets to the other.

More profoundly, though, the extraordinarily tight relationship between the second Trump administration and US tech majors, symbolized by the prominent display of the leaders of X/SpaceX/Starlink, Meta/Facebook, and Google/Alphabet at the new president’s inauguration, signaled that the distinction between Silicon Valley and Washington is disappearing. The paradox is that this will make the US less dominant internationally, even if the opposite was the goal. The power of the Valley was rooted in its capacity for transcending American nationalism. Now that it has full White House backing, the Valley is losing that capacity.

Apart from China, one example of this phenomenon is the Eurostack. The term has an interesting past as it is derivative of the “India stack,” or the subcontinent’s attempt under Narendra Modi to develop domestic digital infrastructure based on control of data, payment systems, and citizen/consumer identity. This is commonly referred to by the acronym DPI (data, payments, identity). When a state can shape and integrate all three of these as the basis for a national digital infrastructure, it can control the nature of its own digital development. Historians of imperialism will savor the irony of the European Union, whose leading members all have imperial pasts of varying extent, looking to the land of the Raj and the Princely States for a model of how to gain control over its digital future. Europe has not often turned to India for geopolitical policy solutions. But that is what it is doing today.

There are counter-currents. For countries like Australia or Taiwan, which find themselves on the frontline of resistance to Chinese digital dominance, joining the US tech sphere of influence makes an immediate sense. The EU is much less sure, and the Trump administration’s indifference to European opinion can only increase its doubts. The US has inadvertently become a driver of digital non-alignment. Assuming India sticks to current policy — and there is every reason to think it will, even if Modi’s own power slips — then the world’s most populous nation and the world’s three largest economies are all pulling in the same direction, which is away from each other.

What of the rest? Consider the UAE. At the end of last year, the UAE’s position, arrived at after long debate and involving considerable discomfort, was to align digitally with what we might have to start calling the America Stack. The symbol of this was the deal last spring between G42, the UAE’s AI-investment flagship, and Microsoft. G42 is run by the UAE’s national security adviser and financed by the state’s sovereign wealth fund, Mubadala. Once the Trump administration’s tech direction became clear, however, G42 pivoted and announced (January 28) that it had become agnostic as to technology.

The demise of global scaling has been gradual over the past decade-plus, but as Ernest Hemingway said of bankruptcy, it can happen “gradually, then suddenly.” Investors now have to pick their way among the India Stack, the China Stack, the America Stack, and (if it happens) the Eurostack. It is unlikely that any invested company will be able to participate, much less thrive, in all four.

The New Geography of Semiconductors

The US CHIPS and Science Act just passed its two-year anniversary, and most coverage focused on the part President Biden emphasized: “America is now on track to produce nearly 30% of the global supply of leading-edge chips by 2032, up from zero only two years ago. … [M]y CHIPS and Science Act is bringing chips manufacturing back to America, strengthening global supply chains.” But strengthening global supply chains and increasing American production are two very different activities. SIG’s view is that the US is playing several distinct games at once: improving US production, isolating the Chinese technology industry, and strengthening certain alliances. The last game is the most interesting one, because it involves US industrial policy as an aspect of US foreign policy.

The US has been somewhat coy so far at naming the countries that will be part of the CHIPS Act-related International Technology and Security Innovation (ITSI) Fund, but the leading candidates are Vietnam, the Philippines, Indonesia, Costa Rica, Panama, Mexico and possibly Kenya. ITSI is administered by the State Department, and Jose Fernandez, undersecretary of state for economic growth, energy and the environment, has said there will be seven core partners without committing to a specific list. He has been promoting the effort as connected to the US wish to secure semiconductor supply chains that avoid China, a window of opportunity he said “may not be here forever.” 

One reason for the vagueness is probably that the Commerce Department has its own priorities, including the Indo-Pacific Economic Framework (IPEF) Agreement Relating to Supply Chain Resilience. IPEF includes three of the possible seven: Indonesia, the Philippines and Vietnam. Then again, Secretary of State Anthony Blinken, before the Americas Partnership for Economic Prosperity, stressed the Western Hemisphere Semiconductor Initiative, which also encompasses three ITSI countries (Mexico, Panama and Costa Rica). 

Some overseas coverage has even mentioned Puerto Rico as being on the CHIPS list, probably confusing the US territory with Costa Rica — although Puerto Rico would also be happy to get involved.

But if inter-departmental rivalry and confused ambitions have made the details a bit murky, the fundamental policy thrust is clear. The US is using the CHIPS and Science Act to both strengthen US semiconductor production and turn a difficult fact — that US domestic wages are high and US engineering talent has better options — into a strategic win. 

The seven countries that have apparently been selected are wildly different. In many ways, the most interesting is Vietnam. It has been making a very strong push to educate engineers, who then become available at a much lower wage than engineers elsewhere: Vietnamese engineers earn half the pay of their Malaysian peers and a sixth of the going rate in Taiwan. This industrial planning is occurring under a Communist government undergoing a leadership transition after 13 years of tight-fisted rule by party general secretary Nguyen Phu Trong, who recently died at age 80. His successor, To Lam, was the tough public-security chief under Trong. His first official trip in his new position will be to China, Vietnam’s largest (and growing) trading partner. Vietnam has a very long history of conflict with China, but of course it also has a shorter but intense history of conflict with the US.

From a distance, it does look odd that the US, hoping to secure its supply chain for vital semiconductor technology, would be relying in part on one Communist state to help weaken another Communist state. 

However, US relations with all seven of the countries on the CHIPS list have had their fraught moments, with the exception of Costa Rica. The technology struggle with China is forcing the US alliance structure into strange new forms. The cold reality is that traditional allies like Japan, South Korea, Australia, Taiwan and others are at demographic plateaus and are priced out of the lower reaches of semiconductor production. A new tier of alliances is being created to deal with this problem. 

The choices being made are often more economic than political — Vietnam’s economy is simply excelling at taking advantage of US-China conflict, as is Mexico’s under left-wing governments that cannot be described as pro-US. However, China was once itself a favored partner despite its ideological coloring — and then, as Xi Jinping’s power and political direction became clear, it was no longer. 

After the New Cold War

To what extent will the U.S.-China struggle take the rest of the world along with it? Recent developments in the technology sector suggest that containment of China has a long way yet to run, regardless of who becomes the next U.S. president. At the same time, China is showing no signs of abandoning its core strategy of using state policy to control citizens at home, build Chinese companies that can crush competition abroad, and exert maximal autarkic control of its domestic market. However, the great success of globalization has been the creation of a global middle class with incomes, educations and expectations all on an unprecedented scale. There is now a generation or two in adulthood that has grown up watching the West destroy itself and slowly abandon the freedom of movement of capital, goods, services and people that was the premise of globalization. This generation, outside the West and (perhaps) China, does not think it is helpless. SIG’s view is that the global generation in its late 20s and early 30s is already pivoting away from attachment to the world of their parents and the disastrous end-game that appears to be their parents’ legacy.

U.S. policy for the technological isolation of China has been steady and focused since about halfway through the first Trump administration. It has expanded in breadth and sophistication under the Biden administration. Technology companies have integrated this into their strategies, giving what began in the government sector strong private momentum. Consider a project with the very Bondian acronym HEIST. It is a private-public-academic partnership now backed by NATO. Its goal is to create ways for Internet traffic to be switched from undersea cables to networked outer-space satellites in the event an ocean cable is disrupted. (Students of Internet history will recall that the Internet itself was developed out of private-public-academic programs for ensuring continuity of communications in the event that land-based systems were disrupted.) HEIST is just one example of how the private and academic sectors are factoring in a long-term tech conflict between the U.S. and China. Another is OpenAI’s decision to clamp down on use by Chinese developers of ChatGPT. China was never on OpenAI’s list of “supported countries and territories,” but the move is nonetheless significant.

Of course, moves like this all call forward responses from China and Chinese companies. China’s GPS alternative, BeiDou, has had this problem set firmly in view for over 20 years. Coverage of the OpenAI decision has emphasized how quickly — measured in days if not hours — Chinese tech companies offered “moving packages” to OpenAI customers on the mainland whose VPN and other outward connections to OpenAI would no longer work. Huawei has retooled itself to deal with the expanding bans on its use overseas. It is too much to say that U.S. tech containment of China has been a good thing for Chinese businesses but it has been a spur, if of a peculiar kind.

The Trump policies on China that Biden kept and developed were guided by people such as Robert Lighthizer (Trump’s trade representative) and Matt Pottinger (Trump’s deputy national security advisor), who are expected to be part of any second Trump administration. There is every reason to anticipate policy consistency, in this particular field, regardless of the victor in November. The same is true in China.

In a real Cold War, this bifurcation between two hostile major powers would extend itself to the rest of the world. There is an element of that today. Germany, for example, after years of U.S. pressure, has decided to take Chinese technology (from Huawei and state-owned ZTE) out of its 5G networks. However, most states and national economies with any choice in the matter have opted either to blend U.S. and Chinese systems or, better yet, to develop their own.

To opt out of a forced choice between major-power antagonists while opting in to the cross-border platforms that are being shaped by that antagonism is a characteristic move for the generation that is now starting its first companies and reaching the lower rungs of government. Chinese autarky and U.S. industrial policy alike have made it clear to the rest of the world that its interests are not of lasting concern to the major powers. At the same time, the spread of middle-class wealth, education and expectations has empowered people around the world to feel they have options. Their politics is shaped by the possibilities for identifying and exercising those options. Ironically, perhaps, for a generation formed by borderless globalization, the chosen venue for exercising those options is not a transnational one but the nation and national or regional economies.

This should not be surprising. Neither the U.S. nor the EU is in any mood to guarantee the sanctity of the global public sphere. China, despite its protestations, is even less globally minded. The fact that addressing global climate change, the signature challenge of the coming generation, is being hobbled by electric-vehicle and solar-panel legislation is truly telling. The major powers that are alone in a position to see through global solutions to global problems are now the very powers making them impossible.

In such a situation, for the world outside the West and China (plus Russia), nationalism and regionalism are the least-worst solutions. The coming generation will be elderly by the time COP75 rolls around and the U.S., EU, Russia and China all bury their many hatchets and rediscover globalism. Meanwhile, away from the current agon, a busy world is identifying problems and designing solutions with no expectation of rising to the universal plane. Globalization has lost its teleology.  But it has created a world in which ambitious people can remain anchored and protected in national economies while also staying closely connected to the world outside, steering their diverse courses with as little reference as possible to great-power conflict.

The Dystopia Fund

The pessimistic worldview reached a new stage this week as the Economist declared that “the liberal international order is falling apart.” SIGnal readers will already be familiar with this idea, but for the Economist to adopt it might reasonably be taken as a sign of the Apocalypse. Meanwhile, the distinguished demographer Nicholas Eberstadt showed at length in Foreign Affairs what a profound demographic crisis East Asia is facing — another SIGnal theme, underlined this week at a desperate press conference given by South Korea’s President Yoon Suk Yeol in which he announced the formation of a Ministry of Low Birth Rate Counter-planning. It was a bad week for optimists, but perhaps the right time to think about a Dystopia Fund.

Macro short-selling is usually done on a national basis, by betting on (or against) currencies. A Dystopia Fund would look to sectors and sub-sectors that would be likely to benefit from long-term negative trends. Humans, even investors, tend to be optimistic, but reality is not. This creates a sort of information disparity that can be exploited.

Last week’s post looked at artificial intelligence in this context. AI, potentially, provides a way for problems of demographic decline to be compensated for by increased productivity within a given national market. The national is important: states large and small have registered that while technology-enabled globalization did indeed bring many tens of millions of people up from poverty, it was driven not by charity but by the extraordinary profits and monopolies that accrued to the companies, overwhelmingly Western, that produced the technology. Now that incomes have grown and capital has accumulated in countries from Saudi Arabia to Singapore, states want to harness technology to grow their own economies and keep wealth in domestic markets. AI promises to do that in that it can increase productivity despite declining populations and, perhaps, poor local education systems.

Other Dystopia Fund targets could be in travel and entertainment for the very old, a growing demographic. Various forms of nostalgia would be investable. Great sums of money in recent decades have gone into market research on people in prime consumer age ranges. Not so much money has gone into understanding elderly consumers. The market-research firm that embraces demographic decline would be worth backing.

An aging population puts transport innovations like self-driving cars into a different light. Most transport systems are only friendly to people who can walk and drive. Companies designing systems for the elderly will find their consumer market steadily increasing. By the same token, tele-medicine, home drug delivery, home medical-testing, and other services for the home-bound will also find growing markets. Working from home, despite its many disadvantages, will provide a way to keep older people active in the job market as the (traditional) working-age population shrinks.

Parag Khanna has led the way in identifying regions that will do relatively well out of climate change — a perfect Dystopia Fund sector. Khanna sees climate-driven emigration as something to be optimized for rather than bemoaned, for the simple reason that it is inevitable. Technologies that can extend the habitability of currently populated areas would be another way to approach the challenges of real estate in a changing climate.

One frequently cited reason to regret the demise of the liberal international order is that challenges like demographic shrinkage and climate change are on such a scale that they would benefit from international coordination. While that is certainly true, liberal internationalism’s vitality has been rooted in optimism, whereas what might be needed now is investment strategies based on a pessimism that accepts the profundity of climate and demographic change and their effects on the status quo. 

Click to subscribe to SIGnal and receive notifications when new posts are available.

AI Family Values

Theories of economic growth do not deal well with demographic decline. How can an economy with ever fewer consumers and producers still grow? Artificial intelligence, looked at a certain way, is an attempt to answer that question. Worries about AI taking jobs are somewhat misplaced. The opposite would probably be worse: shrinking, ageing societies unable to replace lost labor.

AI anxiety usually flourishes in wealthier, industrialized societies. So, for a change, consider Saudi Arabia. It intends to invest $40 billion in AI, which would make it a world leader. Amazon has announced $5.3 billion in AI and data center investments in the Kingdom. Nearby, Microsoft is putting $1.5 billion into a joint venture with Abu Dhabi firm G42, which would also involve the use of Microsoft’s Azure data-center platform. Why such levels of investment?

Yes, there is a geopolitical aspect. Chinese companies dominated the mobile-infrastructure wave of investment in the region in the early part of this century. China’s security-related firms, often banned by the U.S., thrived in the Middle East. Chinese firms used these ties to move into data-center construction. When, sometime late in the Obama administration, the U.S. realized that a great deal of digital real estate was being settled by a hostile power — and that a great deal of American tech capacity was being exiled along globalization’s supply chains — a reaction set in. But even that, by itself, was not enough to get U.S. tech giants investing in the pokey market of data centers in less-developed countries. AI is what made the difference, because AI, in its current configurations, depends on massive, non-latent computing power. The Internet temporarily freed data from physical constraints. AI is helping bring it back to Earth. This happens to coincide with governments’ unquenchable desire to control citizens’ information, in every sense.

But more important is what may be called the gender aspect. To an often underappreciated degree, women are driving technological change. While female workforce-participation rates vary greatly among countries, they have generally been going up for decades, while fertility has gone down. Birth-control technology has been absolutely critical to this change, but its availability and use are social phenomena, driven by choices. The Gulf Arab countries have seen steadily declining fertility rates for years now, along with an easing of prohibitions on the education of women and their participation in the workplace.

In many countries, female workforce participation — greatly enabled by birth control as well as home labor-saving technologies — made productivity and consumer-demand gains possible when male workforce numbers stagnated. In Western countries in particular, there was a sort of internal labor migration from the home to the wage-based workplace. The economic benefits can top out, however — and waged productivity’s gain is fertility’s loss.

Internet-enabled globalization made a second renewal of growth possible. Much like the movement of women into the workforce, globalization put developed-world capital in a position to increase productivity by bringing new labor into the workforce, especially Chinese labor. But this was done at a distance, with laborers outside the home markets of the highly developed economies. Globalization had enormous benefits for even the poorest rich-world consumers, but it did not necessarily make the poor more productive.

AI has the potential to restore bounded, domestic productivity. That makes it intensely attractive to a wide variety of societies, all of which are facing demographic slowdowns or  reverses but have the domestic capital to invest in technological solutions to what is, in the end, a question of power and social cohesion. This is emphatically not about human capital. The Gulf Arab states, despite their wealth, do not have high levels of citizen education such as you find in East Asia or Europe. Saudi Arabia imports human capital, whether laborers or surgeons. About 40% of its population is foreign. Saudi Arabia is a de facto nation of immigrants, but it is not understood as such because it does not expect that 40% of its population to become Saudis. The Saudi state is certainly not worried about AI taking jobs from Saudis. What the Saudi state aims at is achieving the domestic productivity mix that will make it possible for Saudi Arabia to be Saudi, in whatever way the Saudis themselves might define it. Globalization was not especially good for that. Female workforce participation is, as long as it doesn’t undermine “being Saudi.” AI might be even better: in effect, a set of tools for enhancing the productivity of domestic human capital and thereby providing growth without undermining the Saudiness of Saudi, the Japaneseness of Japan, or the Chineseness of China.

AI, then, can be seen as compatible with de-globalization, or more precisely with a decentralization of capital productivity, very much including human capital. It could, in short, enable an increase in domestic production despite a decrease in reproduction.

Click to subscribe to SIGnal and receive notifications when new posts are available.

Networks of Confrontation

When political scientists and policy wonks wrestled with the prospect of a world war that was not, in any conventional sense, winnable — that is, a nuclear war of comprehensive destructiveness — they turned hopefully to “escalation dynamics.” They tried to find a set of reversible steps between chronic conflict and mutual obliteration. It was a way to imagine how to manage the unmanageable. Today, with widespread access to drones and ubiquitous access to the Internet, it is difficult even to define “escalation.” The means for crossing borders, whether in the air or online, have proliferated to such a degree that actors engaged in conflict seem to lose sight of the de-escalation part of the old “escalation ladder.” Having escalated online, they can next escalate with drones, or by activating proxies of one kind or another, or by directing industrial policies toward harming the enemy at one remove. The jumpy, somewhat hysterical, mode of constant irritation of the status quo — constant escalation — was once the signature style of North Korea alone. Now it is worryingly common. 

 There is an argument that the United States was the first mover in this trend toward border-jumping provocation. The U.S., having done far more than any other state to advance the Internet, did tend to treat it as a network for espionage overseas, if not often for conflict. Having pioneered drone technology, which was greatly telecommunications technology, the U.S. made frequent use of it in others’ sovereign territories. With an economy uniquely globalized because of the nearly universal use of the dollar as an exchange currency, the U.S. had another planetary network it could use against its enemies, for example through financial sanctions. And having driven a globalized trading and manufacturing economy, to its own great benefit, the U.S. is now exploiting the resulting global network dependencies by weaponizing industrial and trade policies.

In each of these networks of confrontation, all of which have developed immensely since the end of the Cold War, the U.S. has moved first.

Now so many others are in the game of what the Oxford scholar Lucas Kello, optimistically limiting himself to cyberspace, once labeled “unpeace.” There is so much signaling of malign intent, expressed over cross-border networks, that the foreign-affairs signaling becomes more like noise. Russia’s appetite for information operations directed at undermining U.S. power around the world seems possibly insatiable. One reliable analyst sees Russia’s government as behind a recent infrastructure attack — all executed remotely — on Texas water systems. China increasingly looks to its diaspora as a population physically outside its own borders that is nonetheless expected to show loyalty to the mainland government. Overseas Chinese are monitored and influenced both physically (through embassy and consular staffs) and online. Iran launches an enormous drone attack against Israel and now awaits retaliation, confident that it will not be in the old, pre-1990 form of actual war but in some new “escalatory” move. But escalation without some logic to it is just random warfare. It is unpeace. A lot of aggressive noise without much clear signaling.

For investors, and for everyone else, the challenge is to identify sources and patterns of stability as well as to identify threats. It is not easy. China and Russia are trying to stabilize themselves through a striking combination of patriotism and ethnicity, rallying the tribe of consumers and producers to defend against the external enemy — a method of stabilization that runs quickly into each country’s dependence on external markets for survival. Autarky only feels stable.

The U.S. has the advantage of corporate and entrepreneurial cultures, as well as multinationals, that accept government direction only when they must. Of course the U.S. has many other advantages, and its lack of supervisory power over business has not always been a good thing. But in the present circumstances, when globalized and globalizing networks are both necessary for growth and increasingly dangerous and unpredictable as platforms for political confrontation, the U.S.’s, or perhaps more accurately North America’s, ability as an economy to resist government direction seems to be a distinctive strength. It creates a resilience in unpeace that, one hopes, can survive changes in government. An economy that can thrive under Obama, Trump, and Biden alike is a resilient economy.

Headlines for the IMF annual meeting this week emphasized American economic strength, with growth far outstripping that in the other Group of 7 members. At the same time, the U.S. was at bottom in a poll looking at G7 public confidence in institutions like the military and courts. Politics must dwell on the latter and strive for improvement. But stability comes mainly from the economic side.

Click to subscribe to SIGnal and receive notifications when new posts are available.

The Networks Trap II: Disrupting a U.S.-led System

What could destabilize the emerging geoeconomic framework based on tech networks and alliances? An earlier post (“The Networks Trap,” 6 March 2024) examined how, as one result of the U.S.-China contest, the international system is dividing into two separate and distinct spheres for tech innovation, telecommunications, and military technology. In both the Chinese and American spheres these three sectors are being protected so that American technology will not be in Chinese networks and vice versa. Security alliances in particular, but also political alliances, are increasingly shaped by an emerging type of technological exclusivity. Most of the world would rather not have to choose between one tech ecosystem and another, but the viability of tech non-alignment seems to be weakening. “The Networks Trap” considered this in terms of Southeast and South Asia, where there are flourishing economies, strong domestic tech ecosystems, long-standing wariness of Chinese and American power, and a tradition of non-alignment. The post concluded that tech non-alignment will not be easy to maintain in the region and that the U.S.-led tech ecosystem was likely to prevail in competition with China’s own, even in China’s backyard.

What could disrupt this solidifying geoeconomic pattern?

One potential disruptor is India, which has pursued a technology-development path that avoids dependence on China or the U.S. India learned many lessons from China’s experience building an indigenous tech sector, including that if you welcome foreign venture capital it will come, but once foreign money and expertise have helped you build domestic capacity it might be best to unwelcome foreign capital except under conditions determined by the domestic private sector and the state. (It is perhaps an irony that much of the capital India welcomed and then unwelcomed was from China itself.) India learned from Europe that simply burdening foreign (American) technology with regulations gave no guarantee that domestic entrepreneurs would seize their opportunities. Like most countries, India has neither China’s technocratic relentlessness nor Europe’s high-end purchasing power. After decades of poor results, non-alignment is finally paying off for India. In principle, it could serve as a third-way model to others, such as Nigeria, Indonesia, and Brazil, and this could undermine the U.S.-China-driven tech-telecoms-security alignment currently under way.

However, India’s security worry is not the United States. It is China and its ally Pakistan. India’s security relationship with the U.S. keeps getting tighter, despite very considerable wariness on both sides. What India covets, as the U.S. well knows, is defense technology sharing and joint R&D — as distinct from arms sales. This gives the U.S. leverage to bring India into its networks and even its tech-innovation ecosystem. China’s comparative ability to do this is very weak. India will continue to seek non-alignment but the trend will nonetheless be toward stronger ties to the U.S.

A second potential disruptor of U.S.-China tech bipolarity is South Korean and Japanese dependence on tech exports. South Korea, Japan, and the United States are growing closer in terms of the tech innovation, telecommunications, and security triad. (The Center for a New American Security will release a report on this next week.) The difficulty is that South Korea and Japan, as aging and immigration-resistant societies with thin resource bases, rely on technology exports for growth at just the time when the U.S. is attempting to locate or relocate as much high-tech production to home as it can. This dynamic occurs with European tech powers and Taiwan as well. One partial solution is for these tech economies to relocate production into the U.S., satisfying the U.S. security requirement for production within American sovereign territory while retaining the earnings — in effect, an internationalization of U.S. production that also makes it possible for Japan, South Korea, and the rest to sell into the U.S. market without difficulty. But U.S. preference will still go to U.S. companies, as the extraordinary subsidization of Intel ($8.5 billion in grants and potentially even more in loans) shows. Powers like South Korea, Taiwan, Japan, and Germany cannot hope to be non-aligned in the way India aspires to be, but all these relationships will need to be sensitively managed on all sides, especially the American side. The economic fruits of a U.S.-led tech alliance cannot go disproportionately to the U.S. or its companies.

There are other potential disruptors. China and Chinese tech companies could extend their established practice of selling into markets too unremunerative (much of Africa and Latin America and parts of Asia) or ethically too dodgy (MENA) or both (Central Asia) to appeal to American, Korean, and Japanese tech companies. That would give China a significant advantage. Another possibility is that cybersecurity will become too difficult to maintain across a large number of semi-allied American partners, resulting in resentment of the core — say, the Five Eyes plus Japan, Taiwan, South Korea and some Europeans — by the periphery and political weakness throughout the network.

But the greatest threat is probably an American inability to share either power or wealth. The U.S. is currently engaged in developing defense-production partnerships around the world. The immediate spur has been the Ukraine war: even the Europeans and the Americans together have struggled to keep production at the necessary levels to meet official commitments to Ukraine. Their defense industrial capacity has barely been up to it. The reason is not a lack of military manufacturing capacity per se. Rather it has to do with the downsides involved in the allocation of that capacity to military purposes, which economists generally understand as among the least productive of economic activities: If you manufacture an automobile, it will be used by its purchaser for a variety of purposes that will themselves be economically significant; if you make a tank, its one purpose will be to destroy value, if it is ever used at all. Defense contractors are therefore peculiar animals, protected by states, and enjoying preferential relationships and long-term contracts, in exchange for the preservation and refinement of industrial capacity to provide the arms a state might need to ensure its survival. But in most cases they are a drag on overall productivity, which is why states are often quite happy to buy arms rather than divert resources to producing them.

So for economic as well as political reasons the U.S. does not want to be sole supplier to Ukraine or others. Instead, the U.S. hopes to broaden defense supply chains, seizing what Assistant Secretary of State for Political-Military Affairs Jessica Lewis called “a once-in-a-generation opportunity to transition countries off Russian-origin equipment, improve NATO interoperability, promote transparency and accountability in security sectors, and strengthen our defense industrial capacity.” (Lewis’s Bureau of Political-Military Affairs handles approval of foreign weapons sales.) By “our” Lewis meant European allies, although she went on, in this December 2023 speech, to identify a similar dynamic in Asia. But is it politically or economically possible to internationalize U.S.-allied defense production? Can such an expansion be secure, in cyber and other terms? Will U.S. defense companies be willing to share the risks and profits of production with overseas partners? Will Congress, which determines the contours of defense spending, be willing to let them?

As the international system divides into two separate and distinct spheres for tech innovation, telecommunications, and military technology, one dominated by the U.S. and the other by China, the process has many potential disruptors. The greatest is that the U.S. will be unable to manage this new and untried variety of internationalism.

The Networks Trap

Now that globalization has entered its military or security-led period, what will be the implications for the world’s fastest-growing regions: India and Southeast Asia? These economies have been seeking to refurbish the non-alignment of the Cold War in the hope of benefitting from US-China rivalry. In their different ways, India, Australia, Indonesia, New Zealand, Vietnam, and others have succeeded at turning that rivalry from a zero-sum game into a win-win for them, if not for the US or China. Investment capital has rushed into their economies as a result, with India on track to become the world’s third largest economy. It is early days, however. Neither major players such as India and Australia nor minor ones such as Fiji or Brunei really know what to expect. And none of the state actors involved, even China, truly have the capacity to coordinate defense, commerce, and innovation policies in a harmonious whole. The same is all the more true for a body such as the 14-member Indo-Pacific Economic Framework in the news this week. The potential for policy and economic chaos is very high.

The military connection of the region into an undeclared bloc is not a new development. The Pentagon’s embrace of a strategy built around the China threat has blossomed, under President Biden and Defense Secretary Lloyd Austin, into an extraordinary network of linkages and cooperation agreements throughout Asia. Given bipartisan support for US-China policy in Congress, and the relative policy autonomy of the defense-industrial sector, the marriage of Asian defense strategy and alliance politics seems likely to survive even into a second Trump administration. As America’s domino-theory aggressions in Vietnam and elsewhere in Southeast Asia have not yet been forgotten, it is striking how eager regional powers have been to accept US military leadership in an unwritten SEATO agreement that stretches from Seoul to New Delhi. The world has the Chinese Communist Party to thank for this still rather incongruous development.

As successful militaries have become so dependent on electronics and electronic-network-based technologies, military cooperation and technological capacities have become inseparable. This is one of the two main roots of the global turn toward industrial policy and the repudiation both of globalization—even by national economies that have benefitted most from it—and of the neo-liberal orthodoxies that seemed so immutable not long ago. Political structures such as AUKUS and the Quad are as much tech alliances as security ones. The US is long accustomed to using American technological sophistication and power as a way to cement security alliances. It is equally adept at using security alliances to advance the US defense-tech sector.

At some point, as all the players must realize, there is defense-tech lock-in. And this is where the other main root of the turn toward industrial policy can be found. The strictly economic prosperity of modern nations is dependent both on access to technology and on indigenous capacity to innovate technologically, at least for domestic markets. As so much of new and future technology is dual-use, the overlap of military tech and every other kind of tech has become enormous and seems constantly to grow. Indeed, the term “dual-use” itself, so important a few years ago, already seems anachronistic. When everything seems actually or potentially dual-use, the term doesn’t have any useful work left to do . Semiconductors are simply the most obvious instance. Japan’s decision to revive its own semiconductor industry, at great expense, makes no sense from the old neoliberal globalizing perspective. In the new dispensation, however, it seems a clever policy, at least to the investors who have bid Japanese equities to the highest point in 35 years on the strength of tech companies. The amazing rise in recent years of electronics manufacturing in India — India’s tech weak point since the 1960s — is a similar example of inherently dual-use industrial capacity-building.

Spending state money to build this kind of capacity, which in the old globalization would have seemed absurdly counter-productive in its inefficiency, has begun to seem sensible because it amounts to states investing in their own relative political and economic autonomy. The alternative is to become passengers on networks designed and controlled by others. The question is: Where does this process end? Where are the network boundaries of sovereignty and where are the network boundaries of alliances? The Biden administration’s turn last week toward the potential threats posed by Chinese electronic vehicles is a case in point. Logically, the threat posed by a Chinese  telecommunications network (Huawei, ZTE, China Mobile) is not very different from the threat posed by what the administration identifies as “connected” EVs — or as Commerce Secretary Gina Raimondo has called them, “smartphones on wheels.” The principle extends, perhaps indefinitely, to ships, planes, trains, and any other data-delivery or supply network. The semi-success of an agreement on supply-chain security by the nascent Indo-Pacific Economic Framework should be understood in this context. Even when the supplies as such are not high technology, their transport and distribution networks are. States will continue pushing to secure as much room for maneuver as they can. The brightness of their economic and security futures depends on it. In such a networked environment, however, the scope for non-alignment is shrinking. A later post will consider the factors that could make the emerging geo-economic structure unstable.

The A.I. Scare (I&W)

The A.I. Scare

With the release of ChatGPT in November 2022, the regulation of artificial intelligence suddenly entered the realm of popular politics and media commentary. By May 2023, AI experts were circulating a letter that referred to “mitigating the risk of extinction from AI.” In the same month, the EU Parliament was refining what it called “the first ever rules for Artificial Intelligence.” This wasn’t really true—China got there first—but it did reflect the high level of political energy surrounding the topic. By late June, US President Biden was in San Francisco warning tech leaders about AI’s dangers. The American Enterprise Institute, in line with news-cycle tradition, raised the alarm about excessive regulation.

SIG’s view is that there is both more and less here than meets the eye: “more” in the sense that state regulation of AI is indeed coming on fast; “less” because its effects are not likely to be very dramatic, at least not in the United States.

The Regulatory Wave

China, which early on saw AI as both a strategic technology and a threat to state control of speech and opinion, led the way on AI regulation in 2017 (the “New Generation Artificial Intelligence Plan”) and hoped that its standards would gain international adoption. The powerful Standardization Administration of China (SAC) issued 53 “Guidelines for the Construction of the National New Generation Artificial Intelligence Standard System” in 2020, meeting the deadline set by the 2017 AI Plan. As typically happens in China in moments of political enthusiasm, different bureaucracies began to compete for the new turf and the Communist Party needed to pick some winners and assert its authority. In March 2023, at the Two Congresses, the Party reshuffled bureaucratic authorities and centralized the domestic assertion of AI power, very much in step with the “rectification” of Chinese big-tech power that began in 2021. The extent of Party concern about AI regulation can be measured by the fact that China actually agreed with the United States on some non-binding standards for military AI use in February 2023.

The US, also typically, has been much slower in developing regulations, although it was quick to develop desiderata that don’t seem to have had much real force: Donald Trump’s 2019 Executive Order and a November 2020 Memorandum, followed by Biden’s Blueprint for an AI Bill of Rights in October 2022. The National Institute for Standards and Technology (NIST) took 2 years to consult stakeholders before issuing the “AI Risk Management Framework” in January 2023. The EU was on a similar schedule.

In general, China’s regulatory framework seeks to control information and remind Chinese tech companies that they operate at the pleasure of the Party. The EU aims at identifying and eliminating potential AI harms without constraining innovation. China and the EU both shape much of their efforts with the goal of minimizing dependency on US companies. The US aims at maximizing American innovation and minimizing harms to individual rights.

The 3 efforts reflect very different political cultures, suggesting that they will not be synthesized into broader international standards. In all 3 cases, a principal motivation has been to develop standards that will help each player improve its competitive position against one or both of the other players. It’s hard to see how competition will turn into cooperation any time soon.

So tech regulation is coming, more and more. Unlike at earlier moments of tech revolution, however, there is no free infrastructural (the open Internet) or commercial (unrestrained use of apps) or political (no data sovereignty, no privacy) global platform that AI can build on before it is regulated away. The AI platform is being pre-regulated.

Meanwhile, Back at the Startup

Actual tech regulation, as distinct from the setting out of ideas about things that tech should and shouldn’t do, has traditionally been led by industry. AI will not be much different.

So far, AI innovation has tended to come from smaller companies connected to the open-source community. Broadly, this has been the pattern for tech innovation for decades. It’s possible that AI innovation, too, will follow the pattern that led much of humanity to use the same search engine and a handful of social-media apps: a small company gains a technical advantage, is well run, has the capital to scale its platform without having to generate profits, eats its competitors, and wins big. But the conversation among AI industry leaders today is about whether or not AI innovation will grow based on these same “network effects” rooted in tech, capital, scale, and quantity of data.

The answer might very well be “no.” 

Why? Mainly because AI development teams increasingly turn toward “synthetic data,” that is, curated data sets that are edited to increase the chances of the AI system itself arriving at a desired result—not a specific result, of course, but a result within set parameters, a usable result. This means that the advantage of having huge controllable datasets—which was thought to give China and US big tech significant advantages just a few years ago—is not necessarily that important. It also means that AI development is not a prisoner to the need for scale that so shaped the development of search and social media.

Under these circumstances, AI regulation will be harder to do in any detail because innovators will be small and quick and the use cases for their products hard to predict.

The exception to this small-size advantage is computing power, known in the jargon as “compute”. Large companies with deep pockets have the advantage in compute. That said, compute is itself a product, as Amazon Web Services discovered and proved. And size can be a burden: the biggest companies tend to innovate in ways that take advantage of their size (e.g. lots of compute power) but can also lead them to innovate in ways that don’t matter to the actual market. The history of big-tech failures—Google Wave, Facebook Beacon—is suggestive.