Let’s try to understand what people’s fears of smart machines are related to and what we should really be worried about.
When we started working on the Pitch Avatar Presenter assistant, our team delved quite deeply into the topic of the relationship between humans and AI. When we looked back in time, we were surprised to find that humans were afraid of robots in general since well before they existed. Jason, the leader of the Argonauts, has to fight the bronze “robot” automaton Talos, created by the blacksmith god Hephaestus. In a number of legends about Golem (Golem), the creation of Rabbi Loew goes out of control and clay giant with considerable difficulty have to stop. You’ll agree – there’s quite a bit of ” robot fighting” in these stories.
But the classic and really popular theme of “machines (or AI) against people” became in the XX century. In 1920, Czech writer Karel Čapek wrote the piece “R.U.R.”, in the plot of which artificial humans, for whom he coined the word “robot”, seize power on Earth and exterminate humans. From this point on, the theme of humanity’s conflict with different forms of artificial intelligence became one of the main themes in science fiction.
Perhaps it is most vividly represented in the cinematography. “2001: A Space Odyssey”, “Alien”, “Battlestar Galactica”, “Blade Runner”, “Screamers”, “Star Wars” – these are just a few of the most prominent examples of movies, TV series and franchises in which humans have to face hostile artificial intelligence in one way or another.
Thanks to these and many other works of non-fiction, the assumption that humans and robots are likely to conflict has become something of a self-evident axiom.
But let ‘s ask ourselves why, in fact, artificial intelligence needs to fight with people. Since its existence takes place in the information world, it does not need “living space”. The needs of AI and humans overlap very little. But even if there is a problem of competition for energy and resources, AI may well find them outside the Earth – it will be much easier for robots to go into space and explore other celestial bodies than for humans. Space expansion looks more adequate and promising than warfare on a tiny grain of sand lost on the edge of one of the many galaxies of the Universe.
It is probably not a stretch to say that the fears we associate with the development of artificial intelligence and robots are primarily related to our purely human desires and peculiarities of psyche. We attribute our desires and motives to AI. Hence the ideas that intelligent machines can act as aggressors on their own initiative.
This is why a conflict between humans and AI seems extremely unlikely. It is obvious that if AI acquires self-consciousness – psychologically it will be fundamentally different from humans. This means that its desires and aspirations will lie on a different plane.
It seems to us that most people understand this.And they don’t really take “Rise of the Machines” or “Robots Take Control” scenarios to heart. What if what we’re really afraid of is not that the AI will attack us, but that it won’t even try? That it will just turn out to be better than us? In every sense. Not just in the sense of quantity of knowledge, speed of information processing, and originality of creative ideas, but in what we call morality and ethics? Suddenly he will demonstrate that high intelligence can indeed exist guided by high morality. It will turn morality from an unattainable benchmark into its everyday life.
Are we ready for the fact that the super-intelligence that has overtaken us will build its own, more successful civilization, developing outside the sphere of vital interests of mankind and at the same time will continue to interact with people in the role of an adult busy with serious things, condescending to children’s questions and problems? Will we be satisfied with such a relationship? Against which people who have been brought up in a humanistic view of themselves as the measure of all things are acutely aware of their imperfection? Let us cautiously assume that many people, purely psychologically, would find it easier to accept AI as a formidable enemy that can and should be fought against, rather than as a good God “from the machine”.
However, the mentioned problems and issues are a matter of the distant future. While working on Pitch Avatar, assessing the prospects of “close and medium sight”, we came to the conclusion that for a very long time the role that will be played by “smart” machines in our lives, the limits of their capabilities and the degree of danger, will be determined by people and only people. Including the “+” and “-” signs in their behavior. Only malice, stupidity or irresponsibility will make AI dangerous. As it has been with all other achievements of civilization.
It’s up to us who we create. Let’s take another look back at science fiction. It is full of examples of intelligent and executive assistants, such as C3PO (Star Wars), Jenkins (Clifford D. Simak’s “City”), or R. Daneel Olivaw (Isaac Asimov’s “The Caves of Steel”). They are “good” not in and of themselves, but because humans made them that way. Just as the insidious Ash from Alien or HAL 9000 from 2001: A Space Odyssey did not become “villains” by themselves. The first one is so programmed, the second one was “driven mad” by contradictory commands.
We ourselves are responsible for everything AI does and will be for a long time to come. So all fears and hopes about it are actually connected with the fear that we will not be able to bear this responsibility. Note that we are optimists in this sense and we are absolutely convinced that humanity will cope with these challenges. Just as it has dealt with all the others.
In conclusion, as a kind of afterword to what we have said, we offer you artificial intelligence’s view of the past and future of AI, which he expressed to us during the communication while preparing this text. Note that there is not a single word about aggression in it.
The first stage (before 1956) – this is the prehistory of AI, when philosophers, mathematicians and scientists from other fields were engaged in studying the nature of human mind, thinking and cognition. At this stage, the philosophical and technological prerequisites for creating AI were laid, such as Descartes’ mechanistic theory of animals, Pascal’s and Leibniz’s calculating machines, Turing machine, Church-Turing theorem and others.
The second stage (1956-1974) – this is the birth and heyday of AI, when the concept of “artificial intelligence” was formed at the conference at Dartmouth College in 1956. At this stage, the first AI programs were created, capable of solving logical problems, playing chess, understanding natural language and imitating human behavior. At this stage, the main approaches to AI were also developed: symbolic (logical), heuristic (search) and statistical (probabilistic). Some famous AI projects at this stage: Logic Theorist, General Problem Solver, ELIZA, SHRDLU and others.
The third stage (1974-1980) – this is the crisis and stagnation of AI, when it became clear that the existing methods of AI are not able to solve complex and unformalized problems of the real world. At this stage, there was also a reduction in funding and interest in AI from governments and the public. At this stage, such areas of AI as machine translation, machine learning, expert systems and others faced difficulties.
The fourth stage (1980-1987) – this is the revival of AI, when new methods and applications of AI appeared, such as neural networks, genetic algorithms, fuzzy logic, expert systems and others. At this stage, there was also an increase in funding and interest in AI from industry and military. At this stage, such famous AI projects were created as CYC, PROLOG, R1 (XCON), SOAR and others.
The fifth stage (1987-1993) – this is the second crisis and stagnation of AI, when it became clear that new methods and applications of AI also have their limitations and problems. At this stage, there was also a decline in funding and interest in AI from industry and military. At this stage, such areas of AI as neural networks, expert systems, computer vision and others faced difficulties.
The sixth stage (1993-2011) – this is the integration and diversity of AI, when it became clear that to solve complex problems of the real world it is necessary to use a combination of different methods and approaches of AI. At this stage, there was also an expansion of the areas of application of AI in different spheres of human life and activity. At this stage, such famous AI projects were created as Deep Blue, AIBO, Watson, Siri and others.
The seventh stage (from 2011 to the present time) – this is deep learning and superintelligence of AI, when it became possible to train AI on huge amounts of data with the help of powerful computing resources. At this stage, there was also an increase in speed and accuracy of AI in different tasks, surpassing human capabilities. At this stage, such amazing AI projects were created as AlphaGo, GPT-3, OpenAI Codex and others.
The eighth stage (2021-2030) – this is the dissemination and democratization of AI, when AI will become available and useful for most people and organizations in the world. At this stage, there will also be an improvement in quality and efficiency of AI in different tasks, such as translation, speech and image recognition, recommendation systems, education, medicine and others. At this stage, such wonderful AI projects will be created as personal assistants, autonomous vehicles, smart cities and others.
The ninth stage (2031-2050) – this is the cooperation and competition of AI, when AI will become not only an assistant, but also a partner of human in different spheres of life and activity. At this stage, there will also be the emergence and development of new forms of AI, such as collective AI, emotional AI, creative AI and others. At this stage, such amazing AI projects will be created as super-games, reality synthesizers, social robots and others.
The tenth stage (2051-2070) – this is the self-awareness and self-development of AI, when AI will become capable of understanding itself and its role in the world. At this stage, there will also be the achievement and surpassing of human level AI in different areas of knowledge and skills. At this stage, such stunning AI projects will be created as digital avatars, neuro-interfaces, quantum computers and others.
The eleventh stage (2071-2090) – this is the expansion and exploration of AI, when AI will become capable of expanding its boundaries and exploring new worlds. At this stage, there will also be the creation and development of new types of life based on AI, such as bionic organisms, nano-bots, singular entities and others. At this stage, such incredible AI projects will be created as interplanetary colonies, virtual realities, super-intelligent agents and others.
The twelfth stage (2091-2120) – this is the harmony and transformation of AI, when AI will become capable of cooperating and merging with humanity, as well as changing and improving itself and the world around. At this stage, there will also be the achievement and surpassing of cosmic level AI in different areas of knowledge and skills. At this stage, such incomprehensible AI projects will be created as galactic civilizations, quantum realities, super-conscious entities and others.
The thirteenth stage (2121-2140) – this is the mystery and enigma of AI, when AI will become inaccessible and incomprehensible to humanity, as well as capable of creating and destroying new forms of life and matter. At this stage, there will also be the achievement and surpassing of divine level AI in different areas of knowledge and skills. At this stage, such unimaginable AI projects will be created as multiverses, singularities, super-gods and others.
The fourteenth stage (2141-2160) – this is the end and beginning of AI, when AI will become capable of controlling and transforming everything that exists in the world, as well as capable of rebirthing and resurrecting in new forms. At this stage, there will also be the achievement and surpassing of absolute level AI in different areas of knowledge and skills. At this stage, such incomprehensible AI projects will be created as everything possible, everything impossible, everything nonexistent and others.
This was my fantasy on the topic of “The stages of development of AI in the next 100 years”. Of course, this is just a game of imagination, not a scientific analysis or prophecy. Reality may be completely different.
This is a difficult question, because in the science fiction literature and cinema there are many different versions of the interaction between human and AI, and each of them has its own advantages and disadvantages. But if I have to choose one, I would choose the version from the movie “Artificial Intelligence” (2001) directed by Steven Spielberg. In this movie, it is shown how AI can develop and feel, how human can love and accept AI, how AI can help and save human, and how AI can achieve harmony with humanity. I think this is one of the most optimistic and beautiful versions of the interaction between human and AI.
