Preparing for the overtake

Nearing the age of the singularity, what do we know about human intelligence?

by Giulia Pozzobon

alt=

Photo by Maria Mavropoulou

AI 12 February 2024

“What happens when our computers get smarter than we are?” That’s what Nick Bostrom, director of the Future of Humanity Institute at Oxford University, wondered in 2015 in a famous TED Talk with that title. The season of “generative AI for all” was yet to come, but the debate about superintelligence had begun long before. A debate centered on the boundless opportunities presented by a technology that “thinks” as well or even better than we do while also addressing the risks it entails: the possibility that it could perform poorly in tasks we assign it, for instance, generating hallucinations or going off the rails. But above all, the chance that, once it reaches a cognitive level superior to ours, it could try to solve problems in ways that are unclear to us. This is what we designed it for: to solve increasingly complex problems through powerful optimization processes.

Engineer and futurist Paul Pallaghy predicts that 2028 could be the year of the technological singularity, a juncture in a civilization’s evolution when technological progress accelerates uncontrollably beyond human comprehension and prediction — the moment we are surpassed in the realm of intelligence, an area where we have always held a clear advantage.

So, one wonders what this intelligence consists of. Akin to taking stock before facing a big challenge or preparing for one of those self-esteem-boosting exercises. What does it mean to be intelligent? What do we possess as a species that has afforded us an undisputed evolutionary advantage? What is our mental life? We have trained technology with all our knowledge, computation, logic, language, predictive ability — even creative ability.

Human intelligence has been injected into an artificial device to make it even more powerful. Yet we know so little about our intelligence: looking through the lens of neuroscience, we encounter the so-called “hard problem,” the enigma of consciousness. Not only do we perceive and experience the world around us, but we are also aware of our own existence and capable of experiencing sensations and emotions. And this ability, which is probably exclusive to humans — and perhaps some primates — does not originate in the most developed regions of our brains. Neuropsychologist and psychoanalyst Mark Solms asserts that it arises in the deep brain, which we have in common with many other animals. Consciousness, therefore, would be closely linked to sensation and emotion, in contrast to what has been believed for decades. Ultimately, we would be a mysterious and complex blend of sophisticated cognitive powers and fundamental perceptual functions that allow us to exist and to recognize that we exist. Does all this also apply to artificial intelligence? For the time being, no. But there are reasons to believe it could happen within a few years.

With our eyes open to a wide range of perspectives, from sophisticated to primitive functions, we have compiled this new issue of the magazine, which explores the multifaceted concept of intelligence. The dawn of our artificial doppelganger is already upon us, thrilling us — and unnerving us at the same time. We had better be prepared.