Popular culture has been fascinated by the concept and the possibility of creating artificial intelligence for decades. In literature and cinema there are many stories of intelligent androids and robots, sometimes benign, but in most cases evil and to eliminate us from the equation to become our masters and sovereigns.
The apocalyptic scenarios may be a bit exaggerated, but we cannot deny that technological advances cause us, some, more and more anxiety. An excellent example is the popular techno-horror series Black Mirror in which each chapter addresses a different aspect of our growing relationship with technology and the digital world, usually with chilling conclusions.
Is it possible that advances in artificial intelligence bring us closer and closer to creating superintelligence?
But how close are we to creating true artificial intelligence? Or to be more specific, General Artificial Intelligence (GAI). According to the definition of Shane Legg and Marcus Hutter, intelligence “measures the general ability of an agent to achieve goals in a wide variety of settings.” Under these parameters, there is currently no machine or computer that has this type of intelligence.
Over the years the opinion of the scientific community has been divided into two groups – those who think that we will never create general artificial intelligence and those who think that its creation is imminent. Those in the second group invoke a property of physical laws known as the universality of computation.
According to physicist David Deutsch, this concept implies that
everything that the laws of physics require of a physical object can, in principle, be arbitrarily emulated in detail by some program in a general computer, as long as it is given enough time and memory.
This concept is nothing new and has been explored and analyzed by the mathematician Charles Babbage since the 19th century. Years later the brilliant Alan Turing, father of computer science and precursor of modern computing, would follow this line of thought and in 1950 he would affirm that “a computer program whose repertoire includes all the distinctive attributes of the human brain – feelings, free will, consciousness, and everything – it could be written ”. By the year 2000, Turing claimed, we would have already created these intelligent machines.
Despite this strong scientific backing, advances in the creation of general artificial intelligence have been few and in 2018 we are still very far from fulfilling Turing’s prediction. If science tells us that, in theory, it is possible to create this kind of intelligence, why haven’t we succeeded? What is stopping us?
One possible answer is that we are not faced with a simple programming or computing problem, in Deutsch’s view we are rather faced with a philosophical problem.
One issue that continues to generate strong debate and divided opinions among philosophers is the “difficult problem” of consciousness. The problem is to explain how the brain generates our personal subjective experience and how sensations acquire characteristics. If we still cannot satisfactorily explain these brain mechanisms, then we can’t want to replicate them in a machine.
“I am convinced that the whole problem of developing General Artificial Intelligence is a question of philosophy, not of computational science or neurophysiology; and that the philosophical advance that is essential for its future integration is also a prerequisite for its development in the first place ”.
David Deutsch, Physicist at the University of Oxford
So it seems that to realize the creation of General Artificial Intelligence we must first understand how the brain generates our conscious experience. Although for some this is not a problem. Philosopher Daniel Dennett, for example, denies that there is a “hard problem.” For him what we call consciousness is simply an emergent phenomenon that arises from the connective complexity of the neurons in our brain. Dennett claims that eventually we will fully understand the mechanisms of the brain and end the mystery of consciousness.
But not all his colleagues share this opinion. For some scientists and philosophers, such as Roger Penrose and David Chalmers, the pure electrical and chemical signals in the brain are not enough to explain phenomena such as creativity, intuition, and common sense. There must be something else. The “something else” does not necessarily refer to something spiritual or esoteric; it may be related to quantum coherence (Penrose’s preferred theory) or, as others have mentioned, consciousness may be an inherent property of the universe itself, a theory known as pampsychism.
The reality is that the debate around the “difficult problem” continues to generate controversy and both sides of the coin have good arguments, but not enough. Time continues to advance and it seems that we are still stopped in the same place, although this does not mean that things cannot change drastically from one moment to the next.
What we need, according to Deutsch’s opinion, is just an idea; but of course, it has to be a very good idea.
“It is possible that only one idea is the one that stands between us and the great discovery. But it would have to be one of the best ideas we’ve ever had. “
Deutsch argues that what we need is to have a great idea in the philosophical realm; a theory that explains how creativity arises, or in other words, how the brain creates explanations of our world. In Deutsch’s opinion, creativity is the definitive attribute of an intelligent being, and therefore necessary for the creation of General Artificial Intelligence.
So should we worry about being replaced by super-smart machines? At the moment it seems that we are safe. Although indeed, there are currently computers that can outperform us in very specific tasks, our human creativity remains unique and unmatched for the moment. The creation of General Artificial Intelligence, it seems, will depend on the joint work of various disciplines, not just computer science, and in a certain way it is also necessary to change our concepts of intelligence and knowledge. And maybe then we’ll have that “great idea.”