There is an odd disconnect right now in the public's attitude to AI. On the one hand, the era of hype and fear about the new technology appears to have peaked around 2023 and to have been on a downward slope ever since. We've all gotten used to the presence of chatbots now. And however impressive they are—and I do indeed continue to be wowed that we have built a robot brain that can convincingly mimic human conversation—they also seem to be transforming our lives rather less than we had either hoped or dreaded. In daily life, they often provide little more to us than a variant on the old-fashioned search engine experience—a handy way to gather information from the internet quickly.
Generative AI is in many ways so extraordinary compared to prior generations of information technology that I think we all assumed it would have to change the world. But sometimes, it may in fact be possible to build something amazing that nonetheless has little discernible impact on our daily lives. A few years into this collective social experiment, and jobs have not been systematically displaced by the new tech. The chatbots have proved most adept at drafting unwanted discussion posts or term papers for lazy college students; but any writing or creating that requires going beyond the most generic, mediocre, statistically-average boilerplate still requires a human touch.
Such, at least, is how generative AI now appears to most of us in the mainstream world. The technology has been normalized. And as such, it is all-but forgotten.
But set foot near anyone connected to Silicon Valley, and suddenly, you hear an entirely different narrative. "You normies have no idea what's coming" they say. The new models that are coming out this year have already demonstrated agentic intelligence. They leave the chatbots of 2023 in the dust. They are going to wreak havoc across the entire job market, and the public has not even begun to prepare.
As a recent Wall Street Journal article reports—people in the Bay Area have started to talk about the coming of a "permanent underclass" that will be created through technological displacement. "You'd better start accumulating wealth now," they say, "because this is your last possible chance to get rich." The ark is leaving the station, and only those who own some or all of the handful of AI companies that remain will avoid being swept away in the deluge. Or—maybe not even them. Money itself—many of the proponents of this vision assure us—will shortly become meaningless. So who knows what will ensure wealth and financial security in the brave new world we are entering?
Then, the other day, the Anthropic CEO Dario Amodei offered a more modest—but therefore, more frighteningly plausible—version of the same warnings. He estimated that white collar job displacement from the coming wave of agentic AI models was set to push unemployment into the double digits, while at the same time accelerating economic growth dramatically. Unless government steps in to redistribute some of the gains of this transformation, he said, economic inequality will spiral dramatically—even beyond what it has already.
This picture of the future is a bit less hyperbolic, and a bit easier to believe—it more closely resembles what has happened in prior waves of automation, when technological changes unleashed productivity growth, but at the same time the inherent friction of human societies ensured that large numbers of people were thrown out of work and unable to secure any new opportunities ostensibly being created by the new machines. The "demon of Mechanism" got them—to use Thomas Carlyle's evocative phrase from an earlier stage of the Industrial Revolution—"changing his shape like a very Proteus; and infallibly at every change of shape, oversetting whole multitudes of workmen[.]"
All of this—as the earlier Wall Street Journal article put it—can sound a bit like "science fiction" to us "normies"—who, as pointed out above, continue to be a bit underwhelmed by the actual changes AI has wrought so far.
And indeed, I was just reading H.G. Wells's 1913 book The World Set Free this past week, and I find that his vision of the impact of technological change sounds a great deal like what our current generation of Silicon Valley AI doomers predicts.
Wells's book—which deserves to be better remembered and more widely read than it is (thankfully, it has recently been reissued as part of the "Radium Age" series by the MIT Press)—is astoundingly prescient in multiple ways. Long before World War II—and even a year before World War I broke out—Wells already predicted the eventual development of "atomic bombs" based on the then-recently discovered properties of radium. Not only that, but he also foresaw a coming war with a Prussian-dominated group of "Central Powers" against Britain and her allies. And he predicted that this war would be so catastrophic and total in its effects that it would call for the creation of a new world order.
Wells, writing in 1913, already foresaw the madness and waste that World War I would entail, just a year later. "I kept thinking of the dead Prussian down among the corn," he writes. "[...] Damned foolery! It was damned foolery. [...] From Holland to the Alps this day [...] there must be crouching and lying half and a million of men, trying to inflict irreparable damage upon one another. The thing is idiotic to the pitch of impossibility," he writes.
The war that broke out the year after Wells published these words would prove his point for him beyond any further dispute. The war was an idiotic, catastrophic waste indeed. Wells's "dead Prussian" lying in the corn and moaning out his final agonies makes me think of Thomas Hardy's own prescient words about the wastefulness and stupidity of war—which he likewise published well before World War I was a twinkle in the Kaiser's eye: "quaint and curious war is! / You shoot a fellow down / You'd treat if met where any bar is, / Or help to half-a-crown."
In a 1921 preface that he added to the book, Wells laments that many of his hopes for the war's aftermath had been disappointed. He had foreseen that the looming war would be so insane and pointlessly destructive, that it could only induce a sort of "epidemic of sanity" in reaction. Beholding the spectacle of the Allies's behavior after the war, and the impotence and futility of the new League of Nations, he was forced to concede that these hopes had been premature. The madness continued even beyond the war. Perhaps, he writes, "the disaster had not been vast enough yet or it has not been swift enough to inflict the necessary moral shock and achieve the necessary moral revulsion."
In a sense, he was right. The other prediction he got wrong—he notes in 1921—was that the war came before atomic energy or atomic weaponry were developed. And that perhaps accounts for why the war did not provide quite the object lesson he expected.
Wells ended up being even more prescient than he had realized in 1921—for there would be another world war, less than two decades later. And this one would indeed end in the unparalleled destruction unleashed by atomic weaponry.
And, in a sense, the second world war actually did induce the "epidemic of sanity" Wells had hoped for. The destructive, world-annihilating power of atomic weapons did actually deter overt military confrontation between the great powers for decades to come. The victorious Allies in the second world war—unlike the powers who won the first time around—realized that they had far more to gain from helping their defeated adversaries get back on their feet than from trying to strangle their economies with forced reparations payments and ruinous debts. The United Nations—while far from perfect—proved to be a slightly more effective instrument for securing world peace than its predecessor.
Now, of course, we are living through an era when the madness is all starting up again. Memories of World War II and the lessons it taught have faded with time, and now the "great powers" are invading their neighbors again, violating long-standing treaties, and threatening to annex the sovereign territory of even their own allies. The "epidemic of sanity" has been replaced by an epidemic of madness again—perhaps, alas, the much more common type of epidemic in human history.
Beyond his eerily accurate predictions about the future of war and international collaboration, however—Wells also includes a discussion in the books about the effects of atomic energy becoming available for peacetime purposes. On the one hand, he predicts, the technology will unleash unprecedented abundance. Yet, at the same time—because of an absence of any social foresight or preparation—it causes massive unemployment, poverty, displacement, and deprivation. "If there was a vast development of production there was also a vast destruction of values," he writes—workers in older industries "were being flung out of employment by the superior efficiency of the new machinery."
"The thing had come upon an unprepared humanity;" he adds; "it seemed as though human society was to be smashed by its own magnificent gains. [...] For there had been no foresight of these things. There had been no attempt anywhere even to compute the probable dislocations this flood of inexpensive energy would produce in human affairs. [...] The world was so little governed that the very coming of plenty, in the full tide of an incalculable abundance, when everything necessary to satisfy human needs [...] was already at hand, one has still to tell of hardship, famine, anger, confusion, conflict, and incoherent suffering. There was no scheme for the distribution of this vast new wealth that had at last come within the reach of men[.]"
This sounds indeed a very great deal like the future that Dario Amodei is warning us about.
Indeed, what could be more conspicuous in our own era than the bizarre contrast and discord between the new wealth being created and the unevenness of its distribution—between the extraordinary power of human ingenuity on display in the development of the new AI tools—and the antique barbarism and pig-headed stupidity of our ruling politicians and government officials.
Here we are yet again on the brink—just as Wells predicted—of a possibly exponential increase in human potential and abundance—and yet, the rulers of the Earth are busy squabbling over these gains, showing every sign of trying to hoard the fruits of this abundance and distribute them only according to the absurd and futile and outmoded abstractions of nation, race, ethnicity, borders, etc.
The same years that have brought us an AI revolution, after all—are meanwhile being dominated in our political life by the recrudescence of nationalism, racism, xenophobia, nativism, barbarism, sexism, chauvinism, and every other type of horrible ism.
"Everywhere," as Wells himself predicts, "there were obsolete organizations seizing upon all the new fine things that science was giving to the world, nationalities, all sorts of political bodies, the churches and sects, proprietorship, seizing upon those great powers and limitless possibilities and turning them to evil uses."
Indeed, the AI companies themselves are part of this disgraceful spectacle. While many of them began life with idealistic slogans and protestations on their lips—they have since devolved into flatterers and courtiers of the Trump regime; seeking only to become profiteers of the new wave of racism and stupidity in public life. Palantir's Alex Karp and OpenAI's Sam Altman offer the most starkly and spectacularly debased specimens of this much wider process.
As Wells himself predicts: "a vast multitude of such sharp people were clutching, patenting, pre-empting, monopolizing this or that feature of the new development, seeking to subdue this gigantic winged power to the purposes of their little lusts and avarice."
Of course, it's still possible that we "normies" will be right, and this technology will prove actually less transformative than so many social prophets have prognosticated. As a friend of mine pointed out: many of the Silicon Valley true believers in AI are "effective altruists" who have an essentially religious belief in the coming of the AI superintelligence as a kind of messiah. Their predictions on this score cannot therefore be treated as the dispassionate musings of mere sober technologists.
But suppose for an instant they are right—or right up to a point—and we really are looking at a future where vast new potentialities for wealth creation are about to be unleashed—along with vast new sources of unemployment, displacement, and pauperization on a scale not seen since the days of Heine's Silesian weavers, who were thrown out of work by the arrival of the mechanical loom.
Will we allow what should be a source of collective enrichment to once again become merely a source of widening inequality? Will it merely "ma[ke] the rich richer and the poor less necessary and less free," as Wells puts it?
His admonitions on this point are surely still relevant for us today. He writes of a mass of people left unemployed by the new technology—a "great mass of unskilled cheap labour which the now still cheaper mechanical powers had superseded for evermore. They were being 'scapped'—as horses had been 'scrapped.'"
And in their protests, he writes—"[t]his mute mass, weary footed, rank following rank, protested its persuasion that some of these others must have foreseen these dislocations—that anyhow they ought to have foreseen—and arranged."
Indeed. Let us not let these dislocations from the "demon of Mechanism" once again fall upon humanity unawares. Let us foresee and prepare for what's coming so that the abundance will be shared, not hoarded.
Here's hoping for one more mass outbreak of sanity before it's too late. Heaven knows we're due for it—after the epidemic of madness of the last few years.
No comments:
Post a Comment