Thomas Friedman is Wrong About Artificial Intelligence

Thomas Friedman, an opinion columnist at The New York Times whose work should incite some linguist to invent an antonym for “extraordinaire,” is worried about artificial intelligence (AI). “I’d rather be remembered as the boy who cried wolf when the wolf never came than as someone who got eaten by it. So here goes: ‘Wolf, wolf, wolf, wolf,’” Friedman writes. Even the most charitable spirit must assume that Friedman formulated his conclusions from the reasons stated in last week’s column. If it is so, his argument stands on a foundation roughly as stable as a Jenga tower just prior to the final turn.

Point for point, line for line, almost word for eyeball-singing word, Friedman’s argument stumbles, falls, or wanders off aimlessly into the brambles.

“Here is what my imagination is telling me. It is telling me that the big A.I. companies in America and China are not being candid with us about how little even they understand the power and reach of the models they have developed, like sorcerers, in their labs.” Leave aside for the moment the Chinese. Many American innovators have warned incessantly that their products might cause the sky to fall. Dario Amodei has predicated marketing, public image, and lobbying strategy on it. “There is some chance that is above zero that AI will kill us all,” said Elon Musk in 2023. And so forth. Friedman may be the boy crying wolf—but the supposed wolves, too, are howling that very mantra. It need not be asked whether many technologists fear AI or give voice to those fears—that much is documented in newspapers, blogs, and the archives of television and podcast appearances.

“And [my imagination] is telling me that there is so much at stake—economically, militarily and technologically—and so little trust between Washington and Beijing that it would take a miracle for the United States and China to agree on the kind of deep collaboration needed to defuse the most nightmarish threats posed by A.I.” Friedman is correct on one point: President Donald Trump and Xi Jinping are unlikely in the near future to come to an accord to bind their respective nations and innovators to a single standard. But his imagination, as he terms it, so captured by fears about the effects of a new technology, cannot comprehend the difficulties and dangers of attempting international AI governance in partnership with the Chinese Communist Party. Digital technologies, and particularly AI, embody ideas: ideas about truth and falsehood, morality and immorality, technology’s rightful place in human life, the nature of cognition, etc. The regulation of digital technologies embodies a further set of ideas: ideas about individual liberty and public authority, the means and ends of technological and economic development, the separateness and intermingling of the public and private spheres, etc. One might imagine certain basic technological standards on which America and China could agree—digital technologies and economies are, by nature, transnational, after all. But Americans should not wish for the sort of uniformity of “guardrails” Friedman seems to envision—not even for a day. The common sense and philosophy of the land of liberty are different from those of an authoritarian state, and the concessions that would invariably be necessary to satisfy the latter ought not be tolerated by the former. Americans should not import Chinese governance—especially not governance of the most consequential technology of the modern world.  “Until there is a U.S.-China agreement to ensure that the A.I. agents that the world’s two A.I. superpowers are developing will operate only in ways that are consonant with human flourishing, and be kept that way, I am in the slow-this-thing-down-until-we-are-sure-we-can-control-it camp,” Friedman writes. First, it is doubtful that such an agreement would be consonant with the American tradition; second, it is doubtful that AI regulation which departs from American liberty to accommodate Beijing’s statism would be consonant with human flourishing. The dream of global governance suffers from many defects, but most fundamentally, it forgets that America is unlike any other country in the world—and radically different from many. International agreements, for all their utility in certain instances, carry the risk of de-Americanizing America.

“The last time we trusted the good intentions of the tech bros of Silicon Valley, when they told us to ‘move fast and break things’—regulations be damned—we unleashed social networks that helped fracture our society.” So often, the advocates of safety-ism betray their more fundamental anxiety: a mistrust of technologies that facilitate conversation. “A respected Swiss scientist, Conrad Gessner, might have been the first to raise the alarm about the effects of information overload,” Slate reported in 2010. “In a landmark book, he described how the modern world overwhelmed people with data and that this overabundance was both ‘confusing and harmful’ to the mind.” Gessner’s warning concerned a technology that did much to overturn old authorities and old orders—a technology that, by reducing the cost of disseminating ideas both good and bad and increasing popular access to them, became the tool of political dissenters, religious innovators, and revolutionaries. The 16th-century native of Zurich wrote of the printing press.

Some old lessons remain evergreen; change, technological or otherwise, rarely arrives unalloyed: causing only good or only ill. Social media has not been—and AI will not be—any different. But the rise of the printed word should be regretted no more than that of the digitally carried word. The pessimists of each generation say that generations future ought to be satisfied with the innovations of generations past—enough, they say, no more. To accede to their demands would be an act of theft, an act of negligence, robbing from our children the benefits they might have enjoyed. New technologies often wash over bodies politic like a flood of ice-cold water, eliciting shock, hyperventilation, and panic. But with time, human beings are adept at learning to use technologies well. In the interim, however, in the absence of cool-headedness, fretting about “confusing and harmful” effects and societal “fracture[s]” seems plausible. The Gessners and Friedmans ought not to be followed too quickly and unreflectively.

“Anyone who tells you we can’t afford regulation now because it might slow down innovation is missing three things. First, those who contend that regulation is, by definition, innovation-destroying and market-destroying are ‘just ignorant of history,’ Eric Beinhocker, the executive director of the Institute for New Economic Thinking at the University of Oxford’s Martin School, told me. ‘Good regulations create trust, and trust grows markets. How do you think the airline industry or the pharmaceutical industry grew so big, so fast? Because government regulations created trust among consumers that the planes they fly and the medicines they take are safe.’” Painting the Food and Drug Administration as an icon of prudent regulation ought to disqualify the artist. The transportation industry is another odd example to summon in a column promoting regulation; the sector’s deregulation during the Carter administration proved a tremendous success and a vindication of the market. Hardly anyone, so far as the Taxpayers Production Alliance (TPA) can say, desires the technology sector to be unregulated. Mainline free-market economic theory does not desire any sector to be unregulated. The function of the state as a policeman and enforcer of contracts is virtually unquestioned. Indeed, TPA and the bulk of organizations and commentators favoring permissionless innovation have supported full-throatedly congressionally enacted legislation of the AI industry. The legislation for which we work, however, little resembles the strangulation of liberty and innovation advocated by the pessimists or Friedman’s caricature of “no-regulations” (his phrase) anarchists.

“…A.I. is not just a technology but also, as the computer scientist Craig Mundie has been saying, ‘a new species.’ And like a new species, it is already exhibiting evolutionary properties.” In the phrase of the philosopher Charles Brown, good grief.

As Adam Smith wrote, human beings have the “propensity to truck, barter, and exchange one thing for another”—as well as the propensity to operate upon the world and to improve it, to tinker and to innovate, to invent and to deploy the products of inventiveness. And thank goodness, or else this blog would be painted on the wall of some cave, its author—if he survived childhood—a rude hunter or subsistence farmer, and its argument never distributed to the public. Markets are, as Friedrich Hayek put it, a marvel, and the prosperity they produce, often enabled by technology, also a marvel. Stewarding free innovation well—harnessing innovative energies for good—is one thing; Friedmanite sensationalism and the statist policies that follow it are something else altogether.

Friedman has long chafed at the insistence of our free and open society on not producing the kinds of policies and economic arrangements favored by New York Times columnists (that capacity being another blessing for which we ought to give thanks). “One-party autocracy certainly has its drawbacks. But when it is led by a reasonably enlightened group of people, as China is today, it can also have great advantages,” he argued in 2009, writing that he preferred the Chinese system to the American one. His distrust of AI, which, if freely innovated and freely used, promises to prevent the centralization of power in small self-appointed cadres of the enlightened, should surprise nobody.