Their Beautiful Dark Twisted Fantasies
A review of "More Everything Forever" by Adam Becker
If you need a primer on the crackpot ideas so prevalent in one of America’s critical industries—fantasies that have already warped and distorted public policy in important ways—you’d do well to consult journalist Adam Becker’s recent volume More Everything Forever: AI Overlords, Space Empires, and Silicon Valley’s Crusade to Control the Fate of Humanity. Though he doesn’t dig quite as deeply into the intellectual roots of this millenarian worldview as some readers might like, Becker nonetheless does a considerable service outlining Silicon Valley’s delusions and making clear their incompatibility with the basic laws of physics. It’s a necessary service as well, with the Trump administration actively dismantling the federal government and its scientific research enterprise under the presumption (among other motives) that artificial intelligence will very soon be able to give us all the answers we seek.
But Becker himself strikes a dour, pessimistic note far too often in his narrative, leaving him unable to provide even a basic sketch of his own vision for science and technology in society—a failure that only highlights the desperate need for just such an alternative, now more than ever before.
At its heart, Becker’s narrative illustrates just how quite frankly deranged the ideas that have seized much of Silicon Valley and significant swathes of the government actually are—ideas that bear a greater resemblance to cults or New Age religions than anything else. We don’t need to buy Becker’s general pessimism about the limits imposed by the laws of physics, moreover, to find his case compelling and devastating. The heat death of the universe may no longer be quite as inevitable as we previously thought, for instance, and we may well have been on course to avoid the most likely worst-case climate scenarios before Trump took office for the second time; hard versions of physical entropy and gloomy predictions of climate apocalypse just aren’t needed.
This new Silicon Valley ideology remains “sprawling and ill-defined,” less a unitary suite of beliefs than a constellation of related fantasies from the utopian Singularity and longtermism to interstellar colonization and effective altruism to the existential risk of the imminent emergence of a vaguely conceived artificial general intelligence. It promises salvation and profit all at once, allowing its adherents to ignore pressing problems faced by actually existing human beings and focus instead on realizing the “one-in-a-quintillion chance” that these delusions might miraculously come to pass. Becker calls it an “ideology of technological salvation,” but it’s less an ideology and more a millenarian faith—one that has vast influence over the direction of American society.
As Becker makes exceptionally clear, these articles of faith very quickly run up against the hard realities of pretty fundamental physics and run off the rails of morality when they reach their logical conclusions. Take the notion of the Singularity: the belief that the advent of artificial general intelligence will lead to an inevitable merger of humans and machines that will expand across the universe itself in an “intelligence explosion.” But the technological metric upon which belief in the Singularity rests—Moore’s law—runs up against basic physical limits: engineers and scientists cannot shrink transistors indefinitely; as Becker observes, they can’t go below the 0.2 nanometer size of a silicon atom. As a result, there are also physical limits to the number of transistors that can be fit on a single chip—all of which means that it’ll cost more to achieve additional gains in computing power moving forward. It’s a pretty straightforward case of diminishing returns—and nothing less than a showstopper for the Singularity.
That’s just one aspect of one facet of the Silicon Valley ideology Becker assails in his book. But it’s typical of the poor quality of thinking at play in this intellectual constellation: ignore obvious obstacles rooted in basic science, posit miraculous leaps in technology, and declare the resultant fantasy—or nightmare, as it may be—inevitable. These are not ideas that should be taken seriously, and probably wouldn’t be were it not for the influence they command among and financial backing they receive Silicon Valley’s wealthy venture capitalist class.
That these ideas have seeped into government and public policy ought to be cause for concern. The Trump administration appears set on dynamiting the foundations of American national power and prosperity based, at least in part, on the idea that the artificial general intelligence rapture is right across the corner. Why maintain the infrastructure of scientific discovery—from NASA to the National Science Foundation and NOAA to the National Institutes for Health—when we can just count on AI to figure things out?
More prosaically, the Silicon Valley ideology—in particular fears that current artificial intelligence research poses an existential risk to humanity—prevents us from thinking clearly and pragmatically about the effects, positive and negative, technologies like large language models will have for society. Instead of looking at how LLMs might prove useful for researchers or dissecting the worrying tendencies toward sycophancy and confabulation they display, the U.S. government and political system frets over fanciful notions of “AI safety” and “AI alignment.”
Never mind that the entire ado about AI safety rests on flimsy reasoning sourced to extremely insular online communities, or that LLMs almost certainly will not lead to artificial general intelligence and may well pose an obstacle to it, or that no one has yet offered a clear or compelling definition of AGI itself—or the nettlesome question as to whose values AI should be aligned with in the first place. (Elon Musk, to his perverse credit, seems to think Grok, his own LLM chatbot, should parrot his own far-right views.) As one skeptical tech leader Becker interviewed puts it, “If we’re talking about mythological risks [like AI alignment], then we are completely reframing the problem to be a problem that exists in a fantasy world and its solutions can exist in a fantasy world too.”
The end result, it seems, is to dump money on and devote energy to problems that have little to no grounding in either reality or morality.
Becker’s overall argument would have been even stronger had he delved deeper into the wider threads he only briefly touches upon. His discussion of Russian cosmism and its almost one-to-one parallels with Silicon Valley’s ideology tantalizes readers wanting to know more, for instance, while other historical analogs like the late nineteenth and early twentieth century vogues for theosophy and spiritualism (the latter of which nabbed Sherlock Holmes creator Arthur Conan Doyle hook, line, and sinker) go unmentioned. Still, this is a minor complaint about what’s otherwise a useful primer on the intellectual mania that has Silicon Valley in its grip.
The most glaring weakness in More Everything Forever rests in Becker’s seeming inability to articulate a compelling alternative vision of his own. It’s more than a bit unfair given that Becker sets out to detail and critique Silicon Valley’s fever dreams, not offer something to compete with and replace them. He does his best to pour cold water on these half-baked ideas and does so quite well, but the most he can muster is an impulse to tax billionaires out of existence with a return to the top marginal tax rates seen under President Dwight D. Eisenhower. That’s an idea I’ve become even more sympathetic toward than I was just several months ago, but it’s not a vision for the role of science and technology in human society.
Beyond that proposal, however, Becker’s own worldview too frequently comes across as dour and mirthless. It’s an absence that can’t compete with the insanity on offer from Silicon Valley; something typically beats nothing, no matter how lunatic or impossible that something might actually be. The odd thing here is that it’s not exactly hard to sketch out an alternative based on Becker’s own argument, such as investing much more heavily in robotic solar system explorers and nixing fantasies of Mars colonization—the precise opposite of what the Trump administration has proposed. This failure nonetheless throws the desperate need for an optimistic vision for science and technology into stark relief, one that’s grounded in reality and can provide a compelling alternative to Silicon Valley’s delusions.
But if More Everything Forever helps us shed our illusions about these technological delusions, it’ll have helped start us on our way by freeing us to conjure up our own visions for the future—not outsource them to pseudo-intellectual, drug-addled tech tycoons and their millenarian cults.