The Return of the Magicians

Lately I have found myself writing columns that touch on the rapid advance of artificial intelligence, the mystery of unidentified flying objects haunting American skies and the enthusiasm in certain circles for taking mind-altering substances that yield a feeling of contact with supernatural-seeming entities.

These are very different stories, in a way. The A.I. revolution belongs to the realm of serious and lavishly funded science. The U.F.O. phenomenon hovers on the paranormal fringe. The spiritual dimensions explored by users of drugs like DMT belong primarily to the terrain of psychology and religion — either as manifestations of some sort of Jungian unconscious or else, well, as actual spiritual dimensions. But there is a shared spirit in these stories: the desire to encounter or invent some sort of nonhuman consciousness that might help us toward leaps that we cannot make on our own.

This impulse is an ancient one: The idea that one might bind a djinn, create a golem or manipulate a god or fairy to do your bidding is inscribed deep in the human imagination. Once upon a time this magician’s art seemed like a plausible rival to scientific technique, or a complementary means of mastery over nature; the scientist and the magician were often overlapping figures in the early modern imagination, blurring together in vocations like alchemy and characters like Dr. Faustus. They separated primarily because the scientific method simply worked in a way that magical conjuring did not.

But we are in an era when people talk increasingly about the limits of the scientific endeavor — the impediments to discovering new ideas, the near impossibility, given the laws of physics as we understand them, of ever spreading human civilization beyond our planet or solar system. Meanwhile, the speculations of scientific theorists and philosophers are reaching beyond the very confines of our universe — to an ever-multiplying multiverse whose branches never touch, or an infinite-seeming hall of simulations run by some civilization with godlike capacities relative to ours.

So it is not surprising that our thoughts and efforts might turn back to the magician’s art, in search of powers that might help us escape the limits of our planet. But not simply back to the old magic of spells and incantations. Instead in the U.F.O. fascination and the A.I. enthusiasm and the drug-enabled “psychonaut” explorations, we see attempts to link magic to science, or to deploy science to do magic, using telescopes or chemicals or vast computing powers to discover or create what the old magicians tried to conjure — namely, beings that can enlighten us, elevate us, serve us and usher in the Age of Aquarius, the Singularity or both.

The A.I. project seems to be advancing rapidly, with no speculative leaps required to see its promise. So why lump it in with the dubious and paranormal? Why invoke sorcery to explain a straightforward scientific triumph?

There are good reasons to analyze its efforts in terms of djinns, golems and the like. First, because this is how its own enthusiasts talk.

Here is Scott Aaronson, a computer scientist at the University of Texas, Austin, on his reaction to chatbots:

“An alien has awoken — admittedly, an alien of our own fashioning, a golem, more the embodied spirit of all the words on the internet than a coherent self with independent goals. How could our eyes not pop with eagerness to learn everything this alien has to teach?”

Or consider a recent Wall Street Journal opinion essay by Henry Kissinger, the former U.S. Secretary of State; Eric Schmidt, the former Google chief executive; and Daniel Huttenlocher, the dean of the Massachusetts Institute of Technology’s College of Computing. With the emergent forms of A.I., they argue, we have created an intelligence that can yield answers the way an oracle might: through processes that are invisible to us, permanently beyond our understanding. As such, they argue, the A.I. revolution represents a fundamental break with Enlightenment science, which “was trusted because each step of replicable experimental processes was also tested, hence trusted.” The knowledge granted us by “generative AI” will be far more mysterious; its truth will need to be “justified by entirely different methods, and it may never become similarly absolute.”

And this kind of magical language does not even get into the question of whether A.I. can actually attain consciousness, where the sorcerous aspect of this project is even more explicit. We are telling ourselves that these machines whose workings we don’t fully understand might make the leap to self-awareness if only we keep making their processes more sophisticated, more beyond our ken.

In this sense what we are doing resembles a complex incantation, a calling of spirits from Shakespeare’s “vasty deep.” Build a system that imitates human intelligence, make it talk like a person and answer questions and solve problems through leaps we can’t quite follow, and wait expectantly to see if something infuses itself into the mysterious space where the leaps are happening, summoned by the inviting home that we have made.

Such a summoning is most feared by A.I. alarmists, at present, because the spirit might be disobedient, destructive, a rampaging Skynet bent on our extermination.

But the old stories of the magicians and their bargains, of Faust and his Mephistopheles, suggest that we would be wise to fear apparent obedience as well.

An ancient impulse to seek a nonhuman consciousness.