Much has been said and written about the impact of artificial intelligence on our lives, economy, culture and other areas. This is certainly an issue that will radically change our lives and the world in the future, and the future of AI – the threats and opportunities related to its development – are an area of consideration and action for governments, politicians and the business world.
Paradigm shift?
In recent years, we have certainly witnessed a breakthrough in the development of AI, the likes of which we have not had in decades. According to many experts, the coming years will mean even faster development in this area, encompassing all areas of life. And as they emphasize, it will not be a revolution on a par with other future inventions, but a civilizational breakthrough.
According to the authors of the book “The Era of Artificial Intelligence”, we are dealing with just such a breakthrough. A breakthrough that is only just taking shape.
The publication is authored by three people, two of whom are related to the technology industry – Eric Schmidt, former CEO of Google, and Daniel Huttenlocher, a technology researcher specializing in artificial intelligence, dean of the Schwarzman College of Computing at the influential Massachusetts Institute of Technology (MIT). The third is a man from a different story – Henry A. Kissinger, a diplomat, American secretary of state, advisor to two US presidents, associated with “realistic” international policy based on the balance of power. Kissinger died last year at the age of 100.
If we were to take into account only the first two authors, we might expect a flagship publication on AI in a technological context, but Kissinger introduces a political and philosophical thread to the discussion.
The authors perceive AI as a kind of civilizational breakthrough, comparable to the transition from the medieval world to the modern world, i.e. a fundamental change in thinking about the functioning and perception of the world and man in all areas: “The medieval world had its imago of ideas, its feudal and agrarian patterns, its reverence for the crown and its orientation towards the dizzying heights of cathedral towers. The age of reason had its cogito ergo sum and the search for new horizons, along with this reaffirming human agency within the individual and social concept of destiny. The era of AI has yet to define its guiding principles, its moral ideas and its awareness of aspirations and limitations.”
Your own chess patterns
What made them draw such radical conclusions? It was the progress in the development of “independent” and not just reproducible AI. The latter was based solely on a huge amount of data and is a reflection of human thinking. This is demonstrated by two high-profile breakthroughs. The Alpha Zero chess program developed by Deep Mind, which in 2017 decisively defeated another previously undefeated chess program, Stock Fish. How did this happen and what was Alpha Zero unique about? In the fact that it did not learn to play chess using human patterns, like its predecessors, but, receiving the rules of chess and the goal, i.e. victory, it created its own game patterns through several hours of training consisting in playing with itself. Gary Kasparov, observing the Alpha Zero game, stated that the program played outside human patterns and that people should learn from it.
In 2020, researchers from the Massachusetts Institute of Technology announced the creation of an antibiotic called halicin, a novel drug for bacteria previously resistant to all other drugs. It was created on a similar principle to Alpha Zero. The AI that created the drug by appropriately selecting molecules learned, through training, how to create the final product based on basic preliminary data and databases of molecules. It was to achieve the goals set by the researchers, including an effective drug. However, to this day, the way the AI ”reasoned” and the answer to the question of why it chose this combination from tens of thousands available are not fully known. It is indicated that human work on the antibiotic could take up to several years, and finding the right mixture could prove impossible.
AI also proves superiority on the battlefield. As American simulations have shown, an AI-controlled fighter easily defeats a human pilot in combat in a matter of seconds.
Pushing the boundaries
“AI will seek out deeply hidden patterns and uncover new objective facts, whether making medical diagnoses, detecting early signs of industrial or environmental disasters, or pointing out impending security threats. But in the world of media, politics, discourse, and entertainment, AI will reshape information to fit our preferences, potentially confirming and deepening our biases, and thereby reducing access to objective truth and weakening the consensus around it. So in the Age of AI, human reason will both be strengthened and weakened,” the authors of the book believe.
“In order to push the boundaries of modern knowledge, we can entrust AI with the task of exploring spaces that are inaccessible to us – and upon returning from them, it can provide patterns and predictions that we will not fully understand,” they add.
We can also see the effects of AI in the public sphere; it increasingly influences the outcome of elections, as shown by the recent elections in Pakistan and Indonesia. The imprisoned former prime minister Imran Khann took part in the opposition campaign in Pakistan, and his speech with an AI-generated voice was shared on the Internet and at election meetings. In Indonesia, the presidential election in the first round (after two previous unsuccessful attempts) was won by General Prabowo Subianto, the leader of the local nationalists, accused in the past of brutal actions against protesting students and guerrilla forces. In a broad campaign using AI in modern means of communication, he was presented as a funny teddy bear, a good grandpa, and a cartoon character. And he won huge support among the young electorate.
Global cooperation
However, the key aspect – according to the authors of the book – is the global aspect of the development of artificial intelligence and the rivalry of superpowers in this area. They point to the need for global cooperation on the regulation of AI, comparing the current situation to the Cold War and warning against the possibility of uncontrolled nuclear conflict and other tensions resulting from the decision-making process uncontrolled by humans. Artificial intelligence is increasingly present in this process, and its global management is much more dispersed and fluid today than the management of the nuclear arsenal several decades ago. It is therefore necessary to set the boundaries of AI in the strategic decision-making process and establish a decision-making chain in its relations with humans in key areas.