Artificial Intelligence, often called simply AI, is one of today’s most important technological trends. It is a rapidly-evolving domain with the potential to disrupt virtually all economic sectors. But while the attention is usually focused on its impact on business, AI can also have a major impact on society and politics, both at the domestic and at the international level.
What is driving the development of AI?
Of all the technological advances of our lifetime, AI has the most far-reaching implications. Research in this field has been going on for decades, but in recent years it has become subject of particular interest, attracting considerable investments from both the private and public sector. This development has been driven by the combination of three technological factors. The first is the massive and growing amount of available information that allows Big Data elaboration. The second is the increase of hardware computing power, which enables computers to work on larger quantities of data. The third element is the most important and difficult one: machine learning. Thanks to it, an AI can identify regular patterns and make “experience” upon which it can take decisions. Currently, about 2.2 billion GB of data is generated every day. Tech firms use this data to improve algorithms that mimic human behaviour. Thus, the more users exchange data, the more the service is improved.
Finally, there is a fourth factor: demography. As the population of developed countries shrinks and ages while economic growth slows down, the financial burden of the state mounts. Some advanced economies are already struggling to accommodate these demographic realities; and some are looking to AI to increase productivity and boost economic growth: as a matter of fact, it is estimated that by 2030, AI could add 15.7 trillion USD to the global economy.
Due to the convergence of these factors, the world is on the verge of an “AI Revolution” that will define the 21st century and will radically transform our socio-economic order as the Industrial Revolution did in the past. The applications of AI are countless, and for this reason major firms are investing in this technology and many start-ups are thriving in this sector. For example, between 2012 and 2016 American and Chinese AI firms have received respectively 17.9 and 2.6 billion USD in financing. But corporations are not the only actors that are interested in AI system. States are also involved, to the point that an AI Race has begun between technologically-advanced powers.
States and AI
The first reason why states want to develop AI systems is that this technology can boost their economy and consequently increase their power. AI will not only create new high-income jobs in its own sector, but it would also have a formidable spill over effect in related industries and in the service field; even though there are considerable concerns that many other positions will disappear following the rise of AI. Yet, adopting AI systems will increase productivity in virtually all industries, thanks to faster and more efficient data management.
AI can also assist in complex situations that require the rapid elaboration of large amounts of data; and it can be used to monitor economic, social, political and environmental trends. This has tremendous implications. On the one hand, this allows to take preventive action and avoid dangerous scenarios; but on the other it also means that AI can become an extremely powerful tool enabling governments to control their own citizens and shape their mindset, as well as to interfere in the domestic affairs of other countries. In short, AI is tool to exert power: not much in its coercive “hard” form, but rather in its “soft” version by influencing others via the collection and the spread of valuable information to affect the people’s perception of reality.
Lastly, AI has also impressive military applications. Again, it can have many functions: Command & Control, strategic planning, detection, targeting, cyberwarfare. But it may also take the form of Lethal Autonomous Weapon Systems; or LAWS. These platforms use AI to examine the battlespace and engage targets with various degrees of autonomy. A notable episode dates back to 2016, when an AI powered by a 35 USD Raspberry Pi computer defeated a retired US Air Force colonel in a series of computer-simulated dogfights. Moreover, unlike the colonel, the AI showed no sign of fatigue. What is most important is that this represents a step forward to the ultimate stage of LAWS, which has not yet been achieved, namely “human-out-of-the-loop” systems that they can operate in full autonomy without any human intervention. LAWS are highly controversial due to their ethical implications, and activists are demanding their preventive ban. Yet, their deployment cannot be excluded in the future, at least for platform operating in environments where the risk of unintended civilian casualties is minimal; like air, sea and especially outer space.
This overview already shows the scale at which this technology can serve the state’s interest. As such, it is not surprising that since 2017, many countries have even published their own national strategies on AI. The first has been Canada, followed by Japan, Singapore and China and many other developed economies. These documents differ quite much in terms of length and scope, but it is clear that the cyber domain and AI in the specific are becoming a terrain for great power competition.
In this regard, it is notable that in September 2017 Russia’s President Vladimir Putin stated that “Whoever becomes the leader in this sphere will become the ruler of the world”. Yet, in the AI Race there are two main players that outshine all the others thanks to their huge investments in this sector, their research capabilities and the amount of data they possess: the US and China.
AI bipolarity: The US
Until now, the US remains the leader in this field. Even though it has not yet adopted a unified “AI strategy” document, Washington has clearly shown its interest in this sector. President Trump has taken a free-market stance aimed at supporting American workers, favouring research and innovation, and preserving US primacy in AI. Indeed, this industry is flourishing due to a combination of factors: America is the world’s first economy, it has thriving high-tech clusters hosting large multinational corporations as well as promising start-ups, it holds a very qualified human capital, and both the public and private sectors invest enormous sums of money in AI. But the Pentagon is also actively working on this technology. The latest National Defense Strategy recognizes that China is gaining ground on the United States in the realm of technology; including in the race for AI supremacy. At present, Washington has still the lead, but as the gap narrows it is taking measures to stay ahead: the American leadership is working on a long-term plan that includes inquiries into China’s intellectual property policies and other measures to halt Beijing’s rapid rise. Moreover, in September 2018, the Defense Advanced Research Project Agency announced a plan worth 2 billion USD to develop AI. But the military is also cooperating with private corporations: this public-private interaction is central for the US model, but it has also caused some problems. In June 2018, Google promised not to develop military-related AI and announced its abandonment of a contract with the Department of Defense out of ethical concerns from its employees. Additionally, US multinationals are not fully American firms, since most have parked their assets and corporate profits overseas. In the US and in the West in general, private tech companies have their own goals and governments struggle to shape them via regulation; but this is not the most efficient arrangement and it does not always pay off.
AI bipolarity: China
On its part, China adopts a different approach. Its quest for AI is largely driven by the government, in a model that has been labelled as “innovation mercantilism”. The state coordinates the national AI programme; acts to acquire precious foreign technology and companies; conducts its own research and finances private firms, many of which are often affiliated with the government. China has explicit and ambitious plans for AI: its ultimate aim is to become the world’s “primary” AI centre by 2030, which is a clear challenge to America, which wants to maintain its superiority in this critical field.
In this optic, China has increased its spending on research & development by an annual average of 18% between 2000 and 2015, compared to an annual grow of 4% in the US. (show graphic in source) In other words, Beijing is swiftly catching up, with much more room for further growth; and, due to its sheer size, it has the potential to succeed. With a population of more than 1,3 billion people, China has both the skilled human capital and a formidable pool of data at its disposal; also because its antitrust and privacy laws are weak in comparison to America and the West. Furthermore, in spite of their depiction in the mainstream media, Chinese tech firms operate with large autonomy from the central authorities: their links to it notwithstanding, there are many market incentives pushing them in developing AI. As such, China’s tech sector may even surpass Silicon Valley in terms of market share.
Yet, there are vivid concerns that China will exploit AI as a tool to exert socio-political control on its citizens by establishing a “cyber dictatorship” that evokes Orwell’s 1984 or the gloomiest cyberpunk fiction. While this risk also exists in democracies, this scenario seems much more realistic in China due to its authoritarian government and the huge quantity of cameras installed in its cities.
Russia & the EU
Beside China and the US, the two other main contenders in the global race for AI supremacy are the European Union and Russia. They are aware of the importance of the innovation, yet both lag behind the US and China by considerable margins.
Russia’s problem is that it lacks the necessary capital. Moscow has to selectively invest its limited resources, most of which will go to military applications. For the same reason, like in the past, the Russians are likely to rely on espionage or cyber theft to find shortcuts to AI development.
As far as the EU is concerned, it suffers from fragmented digital market and a complicated regulatory environment. The prominent example is the General Data Protection Regulation, or GDPR; which gives Europeans more control over their personal data and includes fines for unauthorized data use. However, small firms struggle to comply with the high costs required to respect this legislation and are afraid of fines, so the GDPR discourages tech start-ups from growing swiftly.
This is reflected by the EU’s low share of unicorn companies, which are private start-ups that reach a valuation of 1 billion USD. The entirety of the EU has only 29 unicorn firms, while the United States and China hold 139 and 81 unicorn firms respectively. In addition, European tech firms and its experts are the primary target of external recruitment and acquisitions for American and Chinese juggernauts.
The geopolitical consequences of AI
It is evident that AI technology, and the power that comes from it, are largely a prerogative of advanced countries; and this will have geopolitical implications on the global scale.
In spite of the world’s increasing connectivity, states seem determined to affirm and preserve their “cyber sovereignty” by controlling and protecting data at their disposal and by ensuring that AI does not become a menace in this sense. So, instead of adopting a cooperative stance on AI, states tend to behave on the basis of a double trend: the first is “AI nationalisation”, or the tendency to intervene in the economy to direct and control the development of AI; and the second is “AI nationalism”, which refers to using AI to serve the state’s interest and increase its power. This zero-sum logic is mainly driven by the surge of nationalist political forces and great power competition and by the fact that states are aware that AI technologies can be used to exert influence on domestic affairs of other countries and thus pose a threat to national sovereignty. But it is also motivated by the unwillingness of corporations to share their valuable industrial secrets; which is leading to the emergence in various countries of an “information-industrial complex”, often linked to the military, that is gaining more and more economic and political influence
This state-centred approach has potentially destabilizing effects. Not only it fosters competition over deploying AI systems, but also over seizing and exploiting the primary resource needed for its development: data. States will battle in the cyberspace to access and control the largest possible amount of data, as this makes it easier to create effective AI enhances its effectiveness once it is used. China and the US, who already possess huge quantities of data, are poised to hold even larger shares in the future.
This will be closely linked to another phenomenon, namely “cyber neo-colonialism”. Powerful states will use their wealth and technical-scientific superiority to gather precious data from poorer countries in need of advanced technology. These data flows will fuel the development of AI, but at the same time AI systems themselves will also make this kind of exploitation easier; thus creating a self-alimenting circle. This may seem of little relevance, but losing control over data means also that these states will see their sovereignty deteriorate, with important economic and political spill overs that will ultimately leave them even more vulnerable to foreign interferences.
Conclusion: What future with AI?
It is difficult to predict how exactly AI will impact the world’s economic and political order, also because this technology is still in its early stages and its potential is far from being fully deployed, and may even be underestimated. The scenarios are multiple, but one thing is certain: states are competing to develop AI systems, and this technology will have a profound impact on our lives in the coming decades.