[dropcap letter="D"]
uring the nineties, while the Internet revolution was making its way into businesses and homes, little by little we became familiar with the concept of PC, e-mail, the idea of a blog or the changes in habits and new realities that the possibilities of online commerce opened up. Today these are perfectly known realities to the point that not even the least digitally savvy person has difficulty understanding Amazon's business model. Since the late nineties - which will perhaps end up being known as the beginnings of the silicon era - the computational capacity of computers has not stopped growing exponentially: possibilities that illuminate new horizons that only a few years ago were mere hypotheses in science fiction novels. Within this new wave of digital transformation that coincides with the Great Financial Crisis of 2008, what we know as Artificial Intelligence (AI) stands out, one of these new concepts with greater disruptive power and that poses greater challenges in the medium and long term.
AI, whose name is perhaps not the best chosen (personally I am more convinced by talking about "machine intelligence", although it clearly has less marketing), refers to the ability of computers, of machines, to solve increasingly complex problems. So far no problem, at least not related to the frontiers of known knowledge: it is about computers capable of executing increasingly complex sets of programming algorithms that allow giving answers to an ever-increasing number - and which can sometimes seem infinite - of cases and situations. For some time now, Machine Intelligence has been more efficient than human intelligence, for example, when it comes to adding and dividing or calculating the weights of complex structures, optimizing routes and processes, distributing parking spaces on a plot of land, winning in strategy games (even in those with more difficult and elaborate rules), optimizing the setting of air ticket prices or even recognizing faces in airport security cameras, or writing texts, among a very long etcetera.
In all these cases, from a simple sum to facial recognition, these are applications, programs, and algorithms where the universe of possibilities is defined: outside the programmed domain, artificial intelligence demonstrates that it is more artificial than intelligent. The great leap, as Max Tegmark points out in the reference book on the subject Life 3.0 (Taurus), will come when machines are capable of learning on their own, that is, of increasing their own domain of knowledge. Experts in these types of questions, such as the pioneer Ray Kurzweil, identify this moment (which should occur around 2045) as the singularity: from then on, with machines being more intelligent than humans, it will be them, and not us, who will have the capacity to create and manufacture new and better machines.
Tegmark, who in my opinion also has a very reductionist and poor image of what a human being is (a vision, moreover, very common today in Silicon Valley), does strive to build a technological discourse that does not leave out the ethical debate inherent in AI, seeking a ground that allows technical innovation not to conflict with human reality.From this approach also arise, for example, the theses of the historian Yuval Noah Harariexposed in his well-known bookHomo Deus. The main flaw of the book (which I personally believe, despite its good sales, makes no sense) is to reduce complex and multidimensional human intelligence (one only needs to read a bit ofAntonio DamasioorHoward Gardner) to a linear and Cartesian intelligence and, therefore, easily replaceable by a set of algorithms. This analysis leaves aside all ethical-moral or transcendental considerations in the nature of people who are seen as flesh and blood robots. Tegmark's work, in this sense, is somewhat richer in the description of Harari's world of tomorrow, and includes the posing of moral and ethical questions that arise from said diagnosis, although these remain unanswered.
The author, as happens in the bulk of books by other techno-evangelists, sees no limits to artificial intelligence, so it is only a matter of time before it surpasses human intelligence and begins to set its own goals; that is, that, over time, machines will be able to develop their own consciousness as happens in the classic novel by Isaac Asimov I, Robot. The bulk of the text focuses on the description of the main developments achieved in this field and is of great practical interest, identifying with great acuity the challenges that may arise in the medium term in a multitude of fields, starting with the political and freedom terrain or, as I advanced at the beginning, anticipating ethical and moral challenges. Tegmark, who in my opinion also has a very reductionist and poor image of what a human being is (a vision, moreover, very common today in Silicon Valley), does strive to build a technological discourse that does not leave out the ethical debate involved in AI, seeking a terrain that allows technical innovation not to conflict with human reality.
He advocates, in short, for a digital revolution that is not alienated from human affairs and that is not governed solely by economic criteria, but rather maintains a holistic approach to the forces that are driving the Digital Revolution. A proposal that includes a layer of varnish, although not very thick, of renewed humanism also in the field of pure sciences.
Understanding what awaits us in the future involves knowing the technical possibilities of our time, but also, and above all, being good connoisseurs of human nature. That is why I highly recommend reading Tegmark's book (keeping up with the coming world has never been so important), but I encourage you to pair it with a good novel. A Brave New World by Huxley comes to mind.However, and as I pointed out in Harari's case, the comparison made throughout the book between machine intelligence and the nature of human intelligence is very simple and imprecise, which leads to distorted conclusions that may be very far from what truly awaits us in the future. It seems to me a questionable topic, at the very least, to think that flesh-and-blood people with the capacity to feel, create, destroy, move ourselves and move others, to imagine entire universes, among many other absolutely astonishing things that we often don't even understand (it should be remembered that the brain still harbors enormous mysteries for science), can be reduced to lines of code.
Can we reduce to an algorithm the sensations provoked by entering the Sistine Chapel or the central nave of the Sagrada Familia? Can the truth contained in the novels of Proust or Tolstoy be captured? Is the emotion provoked by a musical piece by Beethoven comprehensible to a computer? Morality, ethics, feelings, creative yearning, or the sense of transcendence are part of our physiology as a species; trying to anticipate the future of our species without taking into account our true nature leads us to the same results as those achieved by the neoclassical models of the "homo economicus". Nothing promising. Understanding what awaits us in the future involves knowing the technical possibilities of our time, but also, and above all, being good connoisseurs of human nature. That is why I strongly recommend reading Tegmark's book (keeping up with the coming world has never been so important), but I encourage you to pair it with a good novel. I can think of Brave New World by Huxley, just to give an example.