Comentarios semanales sobre temas de actualidad. Los comentarios se publican todos los lunes, en inglés y en español. Visite los archivos ("Archives") para leer los comentarios anteriores.
Weekly commentaries on current issues. Commentaries are posted every Monday, in English and Spanish. Please, read our previous commentaries in the Archive section.
During the past few days, two stories were announced almost simultaneously: a new religion, Digitalism, whose goal is to worship Artificial Intelligence (AI), was formally launched. And
Alpha Go Zero, an AI developed by Google, learns by itself without human intervention.
Both stories appeared in many media outlets in several languages. Basically, Digitalism is now a real religion, telling us that we, humans, are creating an artificial deity (but very real, it seems) and that we should worship it. And Alpha Go Zero doesnât need humans and, it seems, can self-duplicate.
Letâs see: we are creating a super-intelligent being who will soon realize we are no longer necessary for its purposes. It sounds dangerous. In fact, leading minds, including Stephen Hawking, Elon Musk, and other think so and repeatedly warned about the dangers of AI.
In his book, Superintelligence, philosopher and scientist Nick Bostrom said, âThe creation of a superintelligent being represents a possible means to the extinction of mankind.â
Obviously, there are other more optimistic points of view about the benefits AI will bring to us, humans, in the years ahead. Yet, should we take the risk of waiting until the last minute to discover who is right about AI and our future?
AI keeps progressing and we donât understand it. The situation could be compared (up to a point, of course) to what happened in the United States in the 1860s, when taking pictures was still something relatively new (only two decades old or so). At that time, many people thought photography was simply an ingenious way of deceiving credulous people.
In fact, a New York district attorney, armed with a strong sense of decency and morality, took to court several photographers, charging them with different crimes, from contacting dead people to fraud. The photographers had to explain in court their techniques and eventually the chargers were dropped and photography survived a court challenge.
Of course, 21st century AI is very different from photography techniques in the 19th century. Yet, from a certain point of view, AI seems to be as esoteric and incomprehensible as those early photographs were for those persons who, even looking at them, couldnât understand what they were looking at.
Whatever the case, AI can now teach itself and can duplicate itself. In addition, it AI is global, can access unthinkable quantities of information, and, in fact, it already controls many aspects of our lives.
One more thing: AI knows us better day after day. After all, most social networking sites are just information systems regulated by AI, the AI deciding which posting we will see, in what order, and next to what ads. AI already knows what we like, our weaknesses, and our desires. Even more: AI is not (and perhaps never was) impartial.
What can we do? Letâs find a nice image of a cat or a sunset, or a motivational phrase already repeated a thousand times, and letâs share them, to reaffirm our right not to think about real, imminent challenges.