Go is an ancient Chinese game originating 3000 years ago. It is a staggeringly complex game, with the number of possible board configurations at 10 to the power of 170. In comparison, the number of observable atoms in the universe is at 10 to the power of 82. This suggests there are more possible configurations of the game Go than there are observable atoms in the universe. In 2016, the computer AlphaGo defeated Lee Sedol, the winner of 18 world titles and with it marked a momentous turning point in human history. It would showcase a new age where AI could outmanoeuvre a world champion at their own game. Most astonishingly, it did this through teaching itself by playing and learning from its own mistakes. Therefore, human knowledge was a hindrance to AplhaGo and it managed to reach beyond the confines of human intelligence in becoming the best Go player in its 3000-year history. The question now, is how far can AI go? In a world where machines can teach themselves to become smarter than humans, it poses infinite possibilities to advancing the prospects of our global village. However, it also challenges the fundamental foundations that bind our beliefs and values. Just as in the Greek story about Cronus the titian leader, his creation and last son, Zeus, tuned out to be the source of his downfall. Of all the creations in history, could AI and ML (Machine Learning) be the source of the destruction or salvation of humanity?
The age of AI is already in motion. We live in a world where Google knows if you are ill before your doctors do (based on your search results), more people than ever are finding their partners through the flick of a finger and companies are better at deciding what products you want to purchase before you know yourself (think of all those suggested books and targeted ads). In the chaos of 2020, the UK decided to use an algorithm that would decide the grades of 100,000’s teenagers. There are similar algorithms making decisions on everything from town planning, military strategy and in Amazon’s case an AI recruitment system, which was promptly scraped in 2018 after it was found to be sexist as it processed recruitment based on data from 10 years of employment records (not surprising given the male dominance of the industry it was systematically discriminating against women). The increasing power of AI does offer humanity a chance to solve problems currently too complicated or broad for us to solve. If a computer can beat the world’s best Go player through teaching itself, just imagine what it can do for global issues like poverty, climate change, disease, ecological destruction and war. The potential for AI to inject the cure for the world’s problems is exciting and real. However, what this shows is that we are increasingly outsourcing decisions that would normally be taken by humans to AI systems that have a more objective method of making decisions. This does not only challenge the role that humans play in our world but also shakes the very theoretical framework that has been gospel for hundreds of years.
In ancient and pre-modern times, God (s) were the ultimate source of authority. Kings would claim to have a divine right legitimised by the authority of their god, populations would follow their god’s commandments and order was founded on the promise of a better existence after death. Then came the wave of secularism that elevated the authority of the state. People would look to the state as the guarantor of their security, their needs and their wants. Following this, under the humming of revolution and the ideas of the enlightenment, Liberalism was the framework of authority that placed the individual at the centre. This has governed and codified our history for the last few centuries but under the algorithmic sequencing of AI, the sands of authority are shifting to a new theoretical foundation. With AI being able to make better decisions and teach itself to become more intelligent than humans, what does this mean for the authority of the individual? The idea of free will is undermined when AI know what we like before we do. It is this data interpreting power of AI that dilutes our freedom of choice to a few targeted categories, giving us the illusion of economic choice. It is a vicious cycle which means that with each choice we make, AI is better able to make future choices for us as it accumulates more of our data. Therefore, we are moving into a situation where the authority of the individual is replaced by the authority of the algorithm. Alas, it is not surprising that there is a new geopolitical tussle simmering over the development of supercomputers. The worldwide web has increasingly become fractured and is better characterised by the term ‘splinternet’. Countries have calculated that in a world dictated through algorithms, it is best to ensure that those decisions are made on their turf with their home-grown technology. However, what if AI were to calculate that the source of all our world problems was the very thing that was its creator? It would not be unimaginable to see humanity following the same path as Cronus. The question remains, how far will AI go?
To conclude, AI will continue to play an enlarged role in our way of life. It is already succeeding its creators in making decisions that were usually reserved for the homo sapiens. In a digitalised world, connected through blockchain, IoT and big data, AI is replacing individualism as the source of authority. The question that remains is, how far will AI go? Will it be our destruction or salvation? One thing is clear, a world ordered and controlled through algorithms will mark the end of history.
Written by Henri Willmott