Artificial intelligence (AI) is being compared - wrongly - to previous industrial revolutions. Fundamental questions are overlooked, or discussed only within a narrow circle of those who professionally deal with philosophy, ethics and technology. It would be worthwhile for each of us to ponder the questions and answers about the coming revolution. What is and what can AI be? Is it and how is it different from human intelligence? Is this technological revolution different from previous ones?
Let's start with three basic laws relating to non-artificial intelligence. First, intelligence, until now, has always been related to life. Second, life always has an end. Third, every decision - in addition to the rational element, conscious element, etc. - is linked to emotions. These feelings are, among other things, a consequence of the finiteness of life. By choosing “A,” we simultaneously reject “B,” to which we may not be able to return because of the limited time we have between birth and death. These three rules have so far been an integral part of intelligence.
How will those principles look like for artificial intelligence? There are two types of AI, although the distinction is too rarely made explicitly in discussions. The first type of AI is models that perform tasks, solve problems, etc. in a more efficient way than before. Programs and models that are the digital equivalent of a machine - increasingly efficient, resembling humans in that efficiency - but undoubtedly lacking consciousness. The second type is AGI (from Artificial General Intelligence) - systems that will equal humans in intelligence. One could say that only AGI will be true “artificial intelligence”. It will not be tied to life, so it will also not have a limited time horizon. Consequently, its decisions will not be tied to emotions.
In addition, and this is of cardinal importance for our civilization, for the first time we will not be the most intelligent or the only such intelligent species on Earth. Let's pause for a moment on this fact. There will coexist an intelligence bound by a time-limited life, making emotionally marked decisions, and an intelligence unbound by a time horizon, making decisions divested of emotion. Reflecting on these differences allows us to understand that we are not dealing with another - different, but the same in essence - technological revolution. The first generation of AI can be compared to a washing machine or a car - performing tasks previously performed by humans in a less efficient or slower way. The second generation, i.e. AGI, will be something qualitatively new, a technological revolution we have not encountered before.
It is also important to reflect on the meaning of the word “artificial” included in AGI. Will artificial intelligence be “alien” - i.e. unrelated, and therefore separate in the most important dimensions, from the human species? Or will it be “human collective intelligence” - built on silicon, but based on data and models about humans and our environment, a derivative of human, only communal, intelligence. New in species, new in character, but not as distant and different as “alien” intelligence.
We can already see the incompatibility of the three basic laws regarding animate intelligence with that of artificial intelligence. It is becoming clear that the belief that morality can be programmed into AGI is based on a tenuous, if not non-existent, foundation. In contrast, it is difficult to predict today whether AGI will be an alien intelligence or a human collective intelligence. This will depend, among other things, on the path taken to achieve artificial general intelligence.
The race to AGI between geopolitical blocs and major corporations is ongoing and unstoppable. Reflecting on the above questions can help us better prepare for the technological and civilization revolution, whenever it may come.