Blog: The threat of AI-powered monopolies
You think that Google is a monopoly? With the advent of the AI, tech giants will amass unprecedented power thanks to the peculiarities of AI-powered algorithms. This will force us to completely redefine the current sociopolitical system.
Artificial intelligence and new technologies will undoubtedly bring about tremendous changes, both positive and negative. They will have far-reaching impacts upon our daily lives, our work, security, and values.
On the one hand, thanks to advanced algorithms we will reach previously unimaginable levels of productivity, resulting in enormous benefits to the global economy, as well as drastic improvements in our daily lives with advancements in healthcare, social media, entertainment industry, transportation and a plethora of others fields.
On the other hand, the advent of AI will precipitate massive layoffs and social unrest in the short run, whilst in the long run, AI will challenge the very essence of being humans.
Coupled with issues of abusive use of the AI by authoritarian regimes and unprecedented concentration of power and money at the top, AI is arguably the most dangerous challenge humanity has ever witnessed.
When ordinary people hear of the AI threat to humankind, they usually think of two issues posed by the emergence of AI: massive job losses and total replacement of humans by robots and the future where algorithms will totally dominate us and may even exterminate us. There are, however, other threats that are largely overlooked by people, because media mostly focuses on the first challenge (job losses), whilst Hollywood — on the remote future where robots will supplant humans.
Because of overly focus on these two problems, here I want to closely examine another, often condoned, result of the advent of AI. This is the process that has already begun and is unlikely to be stopped — the formation of AI-powered monopolies.
Already there is a growing discontent both in the media and regulatory bodies over multinational corporations eclipsing nations-states in power and influence. It is has been rightfully noticed that at a time when democratic governments’ capacity to live up to their citizens’ expectations is shrinking, corporations are amassing immense power, not only in terms of money but, most importantly, data.
These companies include so-called MAFIA-G (Microsoft, Apple, Facebook, IBM, Amazon, Google), as well as several Chinese tech giants (Tencent, Alibaba, Baidu, Huawei).
Take interference into elections, for instance. Now, it is Twitter and Facebook, not the US government, who regulate social media by deleting bot, troll and fake accounts and fighting misinformation and propaganda. Another colossus, Google, has stepped up to protect the EU from malignant influence by excluding non-EU sources from buying political ads ahead of the European Parliament elections because the EU lacks a cohesive framework for preventing online meddling.
These are hardly the only examples of multinational corporations designing and enforcing their own public policies. Microsoft recently pledged $500 million to expand the availability of affordable housing in Seattle, which would generally be the job of the US Department of Housing and Urban Development and other public agencies, both state and federal. And at the Paris Peace Forum last November, Microsoft, Google, Facebook, and other tech giants joined 50 governments in signing a new multilateral cybersecurity agreement. Notably absent were the governments of the US, Russia, and China.
When it comes to money, corporations are already more powerful than medium-sized countries. Apple, for instance, when revenue is compared to those of the states, is the 23rd on the ranking of the biggest global revenues, ahead of Belgium, Russia, India, and Mexico.
If we make an alternative list of top 100 economies, there would be 69 corporations and 31 governments, according to the World Bank. More than one-third of world trade is simply transactions between various units of the same corporation rather than between nation-states.
It should come as no surprise that the most breathtaking and monumental projects, such as colonizing of Mars or providing the world with universal Internet access with the help of solar-powered drones or swarms of satellites, come from private companies rather than governments — SpaceX, Facebook, Amazon.
In the past, it was nation-states who launched humans in space and landed them on the Moon. These times are gone. NASA is now considering using private rockets for its new Moon missions in the 2020s, instead of relying on its own resources. And overall, private projects of Mars and Moon landings are years ahead of government plans.
Undoubtedly, the movement of power from governments to corporations will bring about previously unimaginable consequences, whether positive or not is yet to be determined.
However, AI and new technologies are already making Big Tech more powerful than ever before. This is because the AI industry naturally gravitates towards monopolization.
In order to understand why AI tends towards monopolization, we have to delve a little bit deeper into the source of successful AI algorithms.
Nowadays, there are two key things for a powerful AI algorithm: lots of high-quality data and computing power. When data is fed into so-called neural networks, artificial neurons identify patterns within data and yield algorithms. With more data and lots of computing power (needed to perform an enormous amount of calculations when identifying patterns), algorithms become more and more sophisticated and yield better results.
As Kai-Fu Lee, a famous expert on the AI has put it in his book “AI superpowers: China, Silicon Valley, and the new world order”,
the more examples of a given phenomenon a network is exposed to, the more accurately it can pick out patterns and identify things in the real world
Since time immemorial all the way until the Industrial Revolution, the land has been the most important asset, and the more land a person or a government had, the more power it wielded. This is why empires of the past all possessed immense territories, be it Roman, Persian or Maya empires. More land generated more crops, resulting in greater profits and wealth.
Since the Industrial Revolution, land has been diminishing in importance, for factories and machines — the lifeblood of the economy — didn’t need much space. Today, machines and factories are still the mainstays of the global economy, and, notice, the most powerful countries now are not necessarily the biggest ones, for instance, Japan, the UK, Germany, South Korea, because land now plays a secondary role.
In the modern era, however, data will eclipse both land and machines as the most significant asset, and wars will be waged not for the sake of oil or land, but rather for the control of the flow of data. Data will be the primary driver of the economy.
The race to acquire data is already on. Right now, tech giants acquire data by catching our attention with free services and information.
In the short run, this data is used for targeted advertisements that make profits for corporations; however, in the long run, more and more data will eventually perfect the algorithm to the extent that advertisements will not be needed — because algorithms will know us so well that they will make choices for us.
Ordinary humans will find it very difficult to resist this process. At present, people are happy to give away their most valuable asset — their personal data — in exchange for free email services and funny cat videos. It is a bit like African and Native American tribes who unwittingly sold entire countries to European imperialists in exchange for colorful beads and cheap trinkets. If, later on, ordinary people decide to try and block the flow of data, they might find it increasingly difficult, especially as they might come to rely on the network for all their decisions, and even for their healthcare and physical survival. — Y. N. Harari
For better or worse, tech giants will accumulate enormous power with more and more data. Some governments will try to nationalize them and will probably succeed; then the threat of digital dictatorship akin to that in G. Orwell’s “1984” may come true. Many experts are already worried about the abusive use of AI in China and other authoritarian regimes.
I do not think that it would be better if governments would control the data. After all, most people would prefer to give their data to Facebook or Google rather than V. Putin or Xi Jinping.
Nevertheless, it is worth mentioning that above-mentioned scenarios are likely to happen in the long run, probably 10–15 years. In the short run, however, the threat of AI-powered monopolies poses a considerable challenge as well.
As we have found out above, more data means better algorithms. But this creates a positive feedback loop because more data leads to better algorithms and better algorithms attract more customers, who in turn bring more data and amplify algorithms even more.
This will make the process of the concentration of power among a handful of corporations inevitable, and it is unlikely that governments controlling all the data will be a better option.
We the people of the world will likely be forced to completely redefine the current political, social and economic system, as “government tortoise cannot keep up with the technological hare”. The system in which we live now was devised in the age of steam engines and factories. It has not adapted to the age defined by information technologies, as we can see from growing inequality and stagnating wages in both America and Europe. Just as communism emerged as the response to deteriorating living conditions among ordinary workers, it is likely that the advent of AI will create new ideologies and movements of a large scale.
Whether the current system will survive the AI revolution is a matter of time.
- Y. N. Harari “A Brief History of Tomorrow”
- Y. N. Harari “21 lessons for the 21st century”
- K.F. Lee “AI superpowers: China, Silicon Valley, and the new world order”