Liberties There's nothing Artificial about our Intelligence

In the doldrums of last summer, I found myself swept up in a fleeting social-media frenzy. I had thought this could not happen to me again. I had myself written an entire book describing the mechanisms that cause such explosions of irrationality, and counseling readers on how to claw their way out of the naïve and gullible frame of mind that takes claims found on Twitter/X at face value. I had also closed my Twitter account upon concluding “research” for the book. But suddenly I found myself back there, almost unconsciously, disguised behind a new alias account.  The particular frenzy that sucked me in had to do not with artificial intelligence, though there was plenty of that swirling around too, but with the controversial reports of a new substance engineered by South Korean materials scientists, dubbed LK-99. This lab-generated polycrystalline compound was reported to exhibit at least some of the properties of a room-temperature, ambient-pressure superconductor. At present, our superconductors have to be maintained at temperatures and pressures so extreme as to require vast effort, energy, and thus money, to maintain them. But if LK-99 was what some had begun to believe it was, well, this would have been the beginning of a truly enormous technological revolution, with vast, almost unthinkable implications for the global economy and the organization of society. Some compared it to the discovery of the transistor, which inaugurated our current era of telecommunication. Others found even that comparison inadequate. One fellow took to Twitter to declare: “We have discovered fire all over again.”  There was a fascinating scramble to replicate the sketchy results from South Korea, which seem to have been posted precipitously online after a dispute among the members of the lab. Many on social media observed that this was an exciting opportunity for the broader public to watch the scientific method in action. The scientific method, however, as it developed from Francis Bacon to the present day, does not include buzz, or upvoting, or virality, among its mechanisms for arriving at the truth. Yet in the LK-99 fever of the summer of 2023 one would have been hard-pressed to separate the wheat of the findings from the chaff of the buzz.  This became all too clear to me when Sinéad Griffin, a social-media-savvy physicist at the Lawrence Berkeley National Laboratory, ran a simulation proving at least that the Korean results were possible, and linked to her own results along with a GIF file showing Obama’s mic drop at the White House Correspondents’ Dinner in 2016. Griffin’s confidence radiated far, and many found themselves unable to resist it. Even as my own skepticism grew, I remained spellbound. Each night, for a week or so, I skimmed the latest LK-99 results before turning out the lights, closing my eyes, and entertaining visions of a near future of levitating skateboards, quantum computers, ultralight space-elevators to take us to our moon villas, and so much more.  But as an effort to clinch the legacy of presidential administrations and scientific discoveries alike, mic drops seldom age well. (We also saw Obama making the same gesture around the same time on The Tonight Show Starring Jimmy Fallon, upon declaring that Trump would never be president.) In both domains, people are always going to keep right on talking, and if they don’t pick up the mic from the floor this will only be because they have brought their own. In general, we consider this sort of continuity to be vital for the health of both honest science and good democracy. In recent years, however, both science and politics have been warped, sometimes beyond recognition, by the sort of frenzy that we are considering, and of which room-temperature ambient-pressure superconductors, whether they exist or not, whether they can ever exist or not, are perhaps the clearest crystallization yet.  Yes, never-ending contestation is what makes science and politics both work as they should. But this new element — of never-ending posturing by leaders and experts, and of instantaneous camp-choosing and takesmanship by followers and would-be experts — has made it infuriatingly hard to take any reliable measure of where we are really at in the present, and so also of where we are headed.  After my comedown from those few days of LK-99 euphoria, the phenomenon I had just witnessed began to remind me of two others in our recent past, both of which are still sending out ripples in our public debates, and both of which also, perhaps significantly, got delivered to us in the form of acronyms: NFTs on the one hand, and AI on the other.  Now, when I place these together, I do not mean at all to suggest that the science of artificial intelligence is a silly fad, or an abstraction of late capitalism, or something whose future prospects depend on anything similar to what some laconic crypto-bro is willing to pay for a pixel-art image of a Bored Ape. What I mean is that claims of AI’s revolutionary power, either to save our world or to destroy it, must be interpreted in the same way we interpret any other discursive artifact that comes down to us through the filter of the internet. As with effusions about the dawning utopia heralded by LK-99, so, too, warnings of the coming “AI apocalypse” can be assessed with accuracy only when we consider them within the broader context of twenty-first-century apocalypticism: the new habit of interpreting every development, bad or good, dystopian or utopian, as a sign of the end of the world as we know it.  It is hard not to hear the echoes in our present moment of a distant time, when European peasants saw great social transformation around them, along with great suffering, and simply could not conceive a future for humanity on the other side of these transformations. Under such conditions, apocalypse can become not just a prediction, but a fashion. When the rapper and Tamil activist Mathangi Arulpragasam, better known as M.I.A., went to visit