The New York Times is grasping at straws

The New York Times’ reporting on AI is like its reporting in general: partisan, erroneous, sensationalist, and misleading in the extreme.

The tone is apocalyptic, characterizing AI as an imminent threat, much like the Times’ characterization of Iraq’s alleged WMDs. Worse still is the wokeness of it all, the politicization of AI, in which everything — including the survival of the United States — depends on the wisdom and selflessness of the anointed.

Consumer protection, monopoly, economic opportunity, enforcement of existing law, the Constitution — about these things, the Times is silent.

Call it a sign of the times, or a sign that the Times is in decline, because when the supposed newspaper of record soft-pedals the work of rationalists and effective altruists, the result is propaganda with a corporate caption.

That the Times not only “skews left” but skews its coverage in favor of specific policies, that it champions export controls on technology, and that it portrays Anthropic CEO Dario Amodei as a champion of AI safety are reasons the paper reads like a rag with a games section.

Safety is the watchword of the Times’ emphasis on control. All other issues, such as liberty, opportunity, privacy, and democracy, are irrelevant. Or so the Times would have us believe. But first we must believe that AI is a doomsday device, that it is more destructive than the hydrogen bomb and cleaner than the neutron bomb, that it can end mankind not with a bang but a whimper.

Such is the reasoning of rationalists and effective altruists, whose views the Times legitimates.

Whatever else these views are, they are not politically neutral. But because the Times puts “safety” first, and thus sacrifices liberty on the altar of security, there is no room for criticism.

Consumer protection, monopoly, economic opportunity, enforcement of existing law, the Constitution — about these things, the Times is silent. So is Anthropic, which is no accident.

The two, the Times and Anthropic, would use export controls to stifle competition, not promote it. The two would restrict the sale of Nvidia chips to China. Never mind that China continues to buy these chips through intermediaries or on the gray market. Put aside, too, the fact that export controls only serve to goad China into subsidizing the manufacture of its own chips. Or that companies like DeepSeek and Huawei have their own chip ambitions.

RELATED: How to renew American scientific research

BSIP/UIG/Getty Images

Chatham House similarly concludes that hardware export controls alone will not stop China’s advancements in AI. If the controls were effective, policymakers would not have to continually redraw them: banning this or that chip, policing third-country affiliates, responding to smuggling networks, only to authorize new sales anyway. The result is not a technological blockade but a prolonged game of whack-a-mole, which encourages China to be more self-reliant at our expense.

Consider, too, the expense of having the Bureau of Industry and Security involved in this issue in the first place.

Like most government agencies, BIS has a reputation for dysfunction, little productivity, and low morale. Perhaps the New York Times should do a series of stories on BIS instead. Or maybe the paper should stop printing nonsense about “AI safety,” as the term is a catchall for more regulation and higher prices.

A newspaper committed to rigorous journalism would interrogate the interests served by these narratives, confront the failures and contradictions embedded in them, and give serious attention to workers, consumers, civil liberties, democratic accountability, and America’s capacity to out-innovate its rivals.

Instead the Times too frequently amplifies the assumptions of the powerful leftist institutions and woke agendas it should be challenging.