The USA and China: Computers, Conflicts, and Crystal Balls
It is well recognised that over recent years China has developed into a global superpower, one whose GDP is close to rivaling the US, and whose PPP (purchasing power parity) now exceeds the economic dominance of the US. If the US remains, for now, the established dominant superpower of the world, China plays the role of the rising power that threatens a US-centric international order.
China’s reemergence has been a long time coming. Nationally humiliated for hundreds of years by domineering European colonial powers, having lost its Qing ruling dynasty in the early 1900s, China struggled, divided, through the beginning of the 20th century. Left behind by the industrialised world-powers as a third-world country riven by civil war, with much of it under Japanese occupation in World War Two, the country was then thrust into modernity at immeasurable human expense by Mao’s ‘great leap forward’, eventually becoming the manufacturing sweatshop of the world in the early 21st century, before finally being able to forge its current route to global influence and regional predominance. The country’s journey has undoubtedly been an arduous one.
In relation to China, Napoleon once exclaimed,
Let China sleep; when she wakes, she will shake the world.
Napoleon, 1817
It is clear today that the world is most definitely shaking. And more than any other, it is the United States that now feel the tremors as the ground begins to shift beneath it.
The relationship between these two nations, that of a ruling power vs a rising one, is one that Professor Graham Allison has written about at length in his book, Destined for War. In the book, Allison examines the previous historical examples of rising powers displacing dominant ruling powers, and highlights how the last 16 times that this has occurred throughout history, only on 4 occasions has war been avoided, compared to the dozen examples where states have descended into conflict.
This dynamic Allison dubs the Thucydides trap, after the writer Thucydides who extensively recorded for the first time ever an example of war occurring from this relationship. In his History of the Peloponnesian War, the contest at the time was between the ruling might of Sparta, and the growing mercantile power of nearby Athens. Both in that ancient historical example, and in many examples throughout the come centuries, the Thucydidean trap has been sprung despite the actors involved often being determined to avoid war. Many wise heads in these nations knew of the potential damage wrought if they fell into conflict, and yet ended up doing so all the same. We cannot dismiss this fact and assume we are smarter, or have learned the lessons from history, that others did not. Many have felt this way in the past. Many have been proven wrong.
This being said, the book is not fatalistic. Despite the unfavourable odds of war occurring given the historical precedents Allison presents, he also highlights the lessons to be learned from the more peaceful power transitions in history, and how policy makers in both countries can draw important ideas from these. It is a book of diplomacy, not of war. Indeed, there are many obvious reasons as to why both countries would want to avoid conflict. Alongside the tangible human loss of life, economies in our modern world are so interwoven that conflict would spell fiscal disaster, particularly as both states prize their economies above all else.
As a 2016 RAND study noted, in just one year of a severe non-nuclear war predominantly focussed around naval conflicts in the South-China sea and forms of cyber warfare and space satellite targeting, the impact on both countries’ GDPs would be disastrous. The US GDP could sink by 10%, and the Chinese equivalent by up to 35%. Furthermore, we are in the age of nuclear peace and mutually assured destruction. Never before in human history has there been such a threat of near extinction and national suicide from potential conflict, which should further lengthen the odds of war erupting between these global competitors.
Although war will remain an increasingly unlikely risk that neither superpower will want to take, and even if the Thucydidean trap might be successfully escaped from, the two global superpowers will still be in competition with one another for global influence — setting an agenda that works in their interests — and so there will continue to be friction between the US and China played out in every other means other than a ‘hot’ war. In particular it is through geo-economics that the contest for global influence will continue to be fought; the use of economics for achieving geo-political aims.
This is a strategy that both the China and the US are already well-versed in. With China as the largest trading partner of over 130 nations currently, and the US as the undisputed leading economy over the past few decades, they have been able to leverage this fact in achieving their goals. Offering lucrative loans, withdrawing economic support, even simply delaying imports and exports from smaller countries more dependent on them, both competitors know how to more subtly force nations to fall in line behind their aims.
This is but one reason as to why a robust economy with strong growth underpins any form of national power, saying nothing of ensuring domestic growth and success. Thus, as China surpasses the US in economic size, we will begin to see many of the gaps where the US still holds superiority, such as in military capability and living standards begin to narrow. Furthermore, many forecasters remain confident that China’s economic growth is on track to continue, whereas US growth could remain sluggish in comparison.
Yet I believe there is one development that is often overlooked when forecasters peer into their crystal balls at the future however — a development that could simultaneously disrupt or turbocharge the trends being identified by economists and forecasters today. That of AI.
Could AI play a factor in tipping the scales towards an unexpected US resurgence, the strengthening of Chinese dominance in Asia, or increase the odds of future conflict between the superpowers? I believe it could. And I also believe it is often overlooked by forecasters, because there is an ignorance about where ‘AI’ stands as of today, what it is, and what trends will drive its development in future.
The author that best elucidates on this topic is Kai-Fu Lee in his book AI Superpowers: China, Silicon Valley, and the New World Order, and I would urge anyone who wants to understand the state of AI today, and in what direction it is headed, to add this book to their reading list. One of the most important concepts to understand about AI currently is that of deep learning. Deep learning is the most important breakthrough in the past decade in the field, which has “turbocharged the cognitive capabilities of machines” to quote Lee. It applies AI towards a narrow process, and by using a strong algorithm, massive amounts of relevant data, and setting a concrete goal for the computer to reach, an AI can teach itself to recognise deeply buried patterns and correlations in data, which can be used for achieving desired outcomes.
Perhaps the most well-known example of this is in computers playing chess. Whereas before, computers were coded to ‘brute force’ chess, by calculating as many positions as quickly as possible to determine the best position to play. With deep learning however, the computer is coded with the rules of chess, given its desired target (a win), and then left to play itself millions of times (which acts as the relevant data for learning important patterns) where it then discovers which moves are best. Eventually the AI develops moves which may seem counter-intuitive to us as humans, but nonetheless result in victory. When pitted against previous computers that were the best in the world, but that were coded inefficiently in comparison, the deep learning AI consistently outplayed them.
Lee points to Andrew Ng’s simple analogy about AI currently. Ng likens the process of AI deep learning to the harnessing of electricity by Thomas Edison. What we see now, and what media outlets claim are new ‘discoveries’ in AI, are not actually discoveries, but merely the implementation of deep learning to a variety of different fields, just as after Edison harnessed electricity, it was implemented in every field imaginable by entrepreneurs and engineers who had the foresight to do so. We are in an age of implementation, not of discovery, and furthermore, now that the discovery has been made, we are in an age of data, not of expertise.
These two concepts are vital for understanding why China is poised to benefit from the AI revolution even more so than the US. Silicon Valley may have started the AI fire, and certainly has the greatest AI minds in the world, yet it is China who will benefit from its heat. An AI that has had twice as much data to learn from, despite being less well made, will often perform better than a perfectly coded AI that does not have as much ‘experience’, so to speak, processing data.
China can not only harness vast swathes of data more than the US (having more internet users than the US and Europe combined), but also trains more engineers to work on developing the various uses for AI annually, has had a vast government mandate recently for local party officials to offer support for budding AI entrepreneurs compared to the US, and has a cauldron of entrepreneurs forced through immense copycat competition to stay lean and learn to iterate repeatedly in developing the best products for shares of such a vast market.
With this, China is primed to currently benefit more from the AI revolution than the US. PricewaterhouseCoopers estimates AI will add $15.7 trillion to the global GDP by 2030, with China predicted to take home $7 trillion compared to North America’s $3.7trillion. If geo-economics relies upon leveraging the strength of one’s economy over another’s, AI could tip the scales of soft power and influence even further away from the US. Possibly however, this trend could also be disrupted, if the experts of AI learning in the US and Silicon Valley unveil a hitherto unimagined development for AI, unforeseen, which could herald a shift back towards US influence, just as deep learning did to previous decades of AI development. Yet how could AI possibly increase the threat of conflict between these nations?
Deep learning is inherently monopolistic. If an AI works best with a strong algorithm, good engineering, and as much data as possible, how can smaller companies, and indeed, smaller countries, compete with AI products from the US and China that are inherently superior products given their superior engineering and levels of data? Not only will we see an increasing divide between the global haves and have-nots, but also within the economies themselves, we will see increasing inequality between those who own the AI, profiting from it, and those who are replaced and displaced in their jobs by more efficient AI.
There are two pressures that this will cause. Firstly, as countries are divided between those who own the AI and grow wealthy from it, and those who do not, we could see increasing division between those that side with the US and benefit from their technology, and those benefitting from China’s tech instead. An example of this can already be seen in the furore arising over the use of Huawei in the UK and the US, with the cyber privacy concerns that have been raised with their connection to these Chinese government.
As more countries begin to coalesce together under quasi-groupings, this could well increase the tension between the US and China. Just as Greece and Sparta were carried into war from their alliances to smaller city states around them, and just as Europe descended into World War One through their ties of alliances after the assassination of the Archduke Franz Ferdinand in Sarajevo, could we see the US and China brought closer to conflict through their AI-dependent nations and allies?
The second AI based irritant that could cause discomfort between the superpowers is through the marginalisation and the ‘leaving behind’ of more groups in society that are displaced by the AI employment revolution, particularly in the US, beholden as it is to representative democracy. We have already seen populist politicians such as Donald Trump and Nigel Farage garner support from those marginalised groups in society who feel anger and resentment at those in power who they feel have abandoned them. Just as Trump is doing with his trade tariffs on China, he has been able to convince many supporters that the cause of many of their ills can be boiled down to illegal immigrants and foreign enemies, such as China and Iran.
As wealth disparities are set to explode with the advent of AI, as many are gradually shunted into unemployment while others profit fantastically from it, there is a danger that new breeds of populist politicians will increasingly point a finger of blame at foreign enemies, rather than the wealthier classes who will hold more political influence. Of course, the most obvious target of vitriol would be China, given their growing global predominance. Domestic politics could force politicians’ hands into undesirable conflicts for the sake of posturing. If genuine care is not taken to ensure that all of society can profit from the development of AI, rather than a select few, perhaps through a Universal Basic Income (UBI) or otherwise, this could heighten tensions between the two superpowers, who may be more in need of détente than escalation.
When reading predictions about the future, the question of the relationship between the US and China still remains a fascinating one. Undoubtedly it will be the defining global relationship of our lifetimes, just as the Cold War was in years past. I have tried here to bring together two disparate topics that I have not hitherto seen discussed in the same space.
Of course, no topic can be simply divided into its respective parts and then predicted. Any prediction simplifies far too much complexity. But by bringing these topics together, that of diplomacy and AI, in relation to China and the US, I hope to spark some ideas about the links between these two topics and how one might impact the other. As I firmly believe, the best insights are often found at the crossroads between topics, from angles previously unseen, rather than from being ensconced inside one school of thought over another.
I hope this post will inspire you to venture down some avenues of learning you might not have previously walked. I highly recommend both books I have mentioned here — they will be a great place to start on your journey.
OCM
