Adapt and Live

The intelligent questions to ask in the age of artificial intelligence are about speed and care. We've been here before.

“The coming of a tool, then, can be a cultural event of great influence and power. Once that is understood, it is no longer possible to be simpleminded about technological progress.”

-W.B., The Gift of Good Land, p. 105

It’s 1973.

Americans are rushing out to buy microwave ovens. With their saved hours of cooking time, they are enjoying more television. The networks are appealing to younger, urban audiences who have more weight with advertisers. The Beverly Hillbillies, Green Acres and Mayberry R.F.D. have recently been cancelled.

Email and the microprocessor are two years old, and the first mobile phone is being invented. Imperceptibly, the foundations are in place for a coming communications revolution that will warp the speed of information forevermore.

For eight years, Wendell and Tanya Berry have been home on their 12-acre Kentucky farm.

A developer has just failed in his attempt to build on a neighboring property. Using his small salary, Wendell buys 40 acres off the developer and soon realizes he’ll need new equipment to farm them. All he owns is a little rotary tiller, a walking tractor and some implements.

He checks out a used Ford 8N tractor -- an obvious choice. It’s been around for 25 years and is the best-selling individual tractor ever sold in North America. It features the Ferguson System -- a three-point hitch that allows plows, harrows, mowers and scoops to be easily attached and hydraulically controlled, reducing time and effort to operate. Wendell ponders the purchase.

Then Wendell Berry goes and does the most Wendell Berry thing ever. He buys a team of horses instead.

***

Draft horses are the epitome of technological backwardness. So complete was the abandonment of animal power for industrial power that horses (and buggies) quickly came to define slowness to adapt, resistance to progress. Long before the Berrys moved back to Kentucky, the stock of horses and mules on American farms had plummeted 85 percent.

It was the Ford 8N tractor that did in the draft animals, but it too was rapidly becoming obsolete. In fact, everything about farming as it was practiced before World War Two was unraveling. We may have farmed for a time with animals and tools, but motorized machines and mass production became the norm. No one has thought more about the consequences than Wendell Berry. We go to him to understand how corporatism and industrialism came to destroy the green shoots of agrarianism in rural America.

But Wendell noticed something else happening in the second half of the 20th century, though it was harder to see at the time. We were beginning to think about technology as a neutral, values-free force of nature, rather than a human creation of mixed moral significance requiring ethical judgment about how it should be applied. We were beginning our retreat into technological determinism.

In those decades, we were already sliding out of the industrial age into the digital -- with implications far greater than how we would farm. The true significance of the industrial era wasn’t the substitution of physical labor with internal combustion engines. It was a stage-setting for the replacement of human thought itself. The Industrial Age automated physical labor with machines. The Digital Age is beginning to automate our thinking.

***

The first research summit on Artificial Intelligence took place at Dartmouth in 1956. By the 1970s, data processing technologies were automating bookkeeping, typing, and the design of engineering and manufacturing processes. In the early 21st century, the PC and internet automated communication, information sharing and knowledge retrieval. Now we’ve arrived at Generative AI. Machines are creating languages and images that, for millenia and until about one millisecond ago, could only be conjured by human minds.

I know this is only the jillionth post about AI, but I promise you it’s the first to scrutinize the dilemma through the lens of tractors, horses, fields and time. For that we have Wendell Berry to thank, (though no one has), and more specifically his 1978 essay “Horse Drawn Tools and the Doctrine of Labor Saving.” It’s an astonishingly refreshing read in 2025 because in it, Wendell does something that I don’t find in the jillions of AI articles proliferating today. He asks the right questions.

Since we haven’t outsourced all our intelligence-making just yet, I want to lay down a challenge for you.

For the next five minutes, try to drop your efficiency bias against horses and buggies. Imagine that we have systems and norms that help us be selective about technological adoption. Open your mind to the possibility that the road from draft animals to internal combustion tractors to autonomous harvesters might not be the path of progress.

I’m not asking you to choose a side for or against technology. That would be ridiculous. I am a heavy user of artificial intelligence -- am using it to write this very post. What I’m asking instead is to appreciate the questions that Wendell Berry was asking about industrial automation in the 1970s and consider how they might be applied to cognitive automation today. It’s not technology but technological determinism that we’re hitting on this Walk.

***

So it’s 1973 and Wendell is gathering all the antiquated horse-drawn implements he’ll need to get his new team started. The equipment is still around, presumably rusting in heaps and on rafters in his barn and those of his neighbors. He comes to see that the old tools still work well. He cuts his hayfield behind his two horses and, Wendell being Wendell, goes to inspect other nearby fields, all mowed with tractors.

I can say unhesitatingly that, though the tractors do faster work, they do not do it better... Through the development of the standard horse-drawn equipment, quality and speed increased together; after that, the principle increase has been in speed. Moreover, as speed has increased, care has tended to decline.

This then is the seminal question to be asked about a new tool or technology: does it enable the work to be done faster, or better, or both? Wendell observed that tractors allowed farmers to work faster, which meant they could do more work within those fixed hours of daylight. But this didn’t mean that they could work better. “And there comes a point, as we know,” he says, “when more begins to imply worse.”

Faster and slower are easily measured. Better and worse are not. It’s a little appreciated feature our digital zeitgeist that we’ve become totally dominated by the rules of data, by a language that can only be written in one’s and zeroes. Questions of quantity, no matter how difficult, are at least answerable. The questions and answers share a language. Quantity is King because it is the core feature of computation, computers, and their millions of marvelous applications. But Quality? Quality is the stranger, the pariah.

So new tools don’t necessarily allow work to be done better, only faster, and figuring out “better” is way harder than measuring speed. But isn’t there tremendous value in speed? Isn’t having more time, or less labor, the very point of our technological pursuits?

Maybe. But here we come to the next of Wendell’s questions. Why haven’t we asked for what purpose and for whom is our labor being saved? By not asking this question with the advent of every new technology, we leave it to the experts and the market to make that determination for us. What becomes of those saved hours? Can we use them in the same places where they were saved? Are they deposited in the savers life account, to be enjoyed in the best way they know how? “To widen,” as Wendell puts it, “the margins of leisure, pleasure and community life.”

Hardly. The impact of labor-saving technology has been more capital for the capitalists, and for the workers -- more wage work to pay off debt; retraining for jobs that haven’t yet been automated; idleness; drug use and, too often, despair. More time has not meant more value for the worker whose time has been saved. Just the opposite in fact.

It appears that we abandoned ourselves unquestioningly to a course of technological evolution, which would value the development of machines far above the development of people.

Technological determinism has the same problem as capitalism. Their sole function is to feed themselves. Capitalism exists to expand capital (we call this growth, far too casually), and technology, through data accumulation, will devour every last thing that can be digitized. But these are not systems operating under natural law. Like our economy, they are managed by us. Unless they aren’t.

Artificial intelligence isn’t frightening because it can solve problems by finding patterns in data; it’s frightening because we know ourselves too well. Our track record of managing dehumanizing systems with guardrails and safety nets is lousy.

Maybe it’s a curse -- being the tool-making animal. We don’t know how to stop ourselves, or even pause, when doing so could mean the difference between health and harm. We’ve conflated speed and ease with quality and progress, and we’ve left out the most essential criteria of all, which Wendell never does: what does it mean for people? Ironically, I suppose, this presents us with a binary choice. Every time a new tool appears, we have to ask: is it socially benign, or not?

The coming of a tool then, is not just a cultural event; it is also an historical crossroad -- a point at which people must choose between two possibilities: to become more intensive or more extensive; to use the tool for quality or for quantity, for care or for speed.

We need to starting choosing technologies for how they help us extend care and not just save or fritter time. For their contributions to health and not just pleasure. Our experiments with social media are still playing out (badly). That should be clear to see because they are recent and impacting all of us today. Can we finally see how industrializing our agricultural system was also a treacherous set of decisions for our health? How the “adapt or die” ethos of ex-Secretary of Agriculture Earl Butz, and every disrupter since, contains within it, a declaration of harm?

Contrary to what all his critics say, Wendell Berry isn’t hoping we return to earlier times. He isn’t arguing for animal power, or against computers, or for collectives and co-ops, though he’d surely be in favor of more of all that. In those sweet-bright days he writes about in his poetry and fiction, we were already bowing down to the time-saving gods of technology.

He doesn’t want to stop time. What he wants is for us to bring our human intelligence to the challenge of adaptation. We must stop adapting our lives and systems to accomodate technologies that harm our communities and souls. It’s just the other way round. We must adapt our machines, tools, algorithms, and large language models to serve our human needs “as our history, our heritage, and our most generous hopes have defined them.”