AI and Leisure Polarization

You may have noticed that we’ve been pretty quiet here at the Savage Collective. I’ve been traveling—both for work and vacation—and, to be honest, dealing with some health issues that made it nearly impossible to write. Brandon’s been just as swamped, with his own travel and professional obligations. The Substack is a side hustle for both of us. But we’re both committed to getting our writing back on track. Also, remember that we’ll be hosting a gathering in Ligonier, PA November 7-8. It is going to be great with lots of excellent writers. So, we hope to see many of you there.

In the meantime, as we get our writing back off the ground, I wanted to share a few brief reflections on some recent pieces Tyler Cowen has published in The Free Press about generative AI.

The Chess Players by Honoré Daumier

Let me start with this: I’m a big fan of Tyler. He is an economist at George Mason and, when I was in graduate school, I read his blog —Marginal Revolution—religiously like many other econ-minded PhD students. His podcast, Conversations with Tyler, is among the best out there. He’s exceptionally smart and truly one of the most well-read people on the planet. While I don’t always agree with his libertarian leanings, I find him to be deeply curious and largely guided by goodwill.

That said, I do think Tyler has one significant blind spot when it comes to technology—particularly in how it affects people who aren’t elite or unusually gifted.

Back in April 2024, Tyler hosted Jonathan Haidt to discuss Haidt’s new book The Anxious Generation on his podcast. What stood out most from that conversation was how worked up Haidt became—so much so that he even cursed on the podcast, which genuinely shocked me. Jonathan has always struck me as level-headed and measured. His frustration was clearly provoked by Tyler’s persistent refusal to concede what seems obvious to nearly everyone: smartphones and social media are, on balance, harmful to kids.

Tyler pushed back on several fronts, but the most remarkable to me was his fixation on how social media helps the very high-achieving students he mentors connect with similarly exceptional peers. In other words, his defense of the technology was centered on the benefits for a tiny sliver of extraordinarily talented individuals.

And that, I think, is the heart of the issue. Tyler often evaluates technology through the lens of how it further empowers the gifted, without giving enough thought to how it shapes the lives of the vast majority of “normies” who live outside those rarefied circles.

He takes a similar approach in his recent Free Press essay, “Does AI Make Us Stupid?” There, he critiques a study by MIT professors suggesting that while AI might make tasks easier, the downstream effects could be perilous for our long-term ability to think. The study essentially argues that tools like ChatGPT reduce the cognitive effort we put into tasks, potentially leading to the atrophy of key intellectual skills. It’s a concern I raised in a recent piece of my own. My primary concern when it comes to AI is that we will lose central human capacities such as reading and writing.

However, Tyler does raise an important counterpoint. He asks: If we’re using less cognitive capacity on certain AI-assisted tasks, what are we doing with the time we’ve gained? It’s a question I often pose to my own students. Efficiency, I remind them, is not a basic good but an instrumental one. We have to ask: Efficient toward what end?

Tyler essentially argues that we will be able to turn more of our attention to the things in life that are most meaningful to us. He argues particularly that we will gain more leisure. Generative AI will relieve us of the drudgery of high-load cognitive tasks, freeing up time for good leisure. He says this in a separate piece called “AI Will Change What It Is to Be Human. Are We Ready?”

To live a happy and fulfilling life without work, one needs meaningful pursuits, social connection, and physical movement. We don’t expect any of those needs to go away in a world of AI ascendance. Indeed, those needs will likely become more intense.

He may be right, and this could be a net good especially if we understand leisure as not entertainment but instead in the way that Josef Pieper talked about leisure—as the contemplative time in which we engage with what is good, true, and beautiful.

Tyler gives his own example of how he’s using this reclaimed time with ChatGPT:

I am doing [this] on my current travels in Spain and France. [People] can ask [ChatGPT] to give the history and background on each church, as I also am doing. I learned, for instance, how the church in Reims, France, brought in Marc Chagall and later the German artist Imi Knoebel to help spruce up the presentation in previously damaged windows.

Honestly, I don’t even need to comment on how out of touch this feels for most people, who don’t have Tyler’s money—or his intellect.

The problem is that the time digital technology gives with one hand, it often takes away with the other—flooding our leisure with cat videos, pornography, “funny fails,” and endless ads. I’ve done enough teaching and traveling with Gen Z to know that most are not using time saved by digital tools to read the Bible or study Peronism.

Since 1980, the U.S. has undergone rapid deindustrialization, shifting from manufacturing to service-sector jobs in education, healthcare, finance, and beyond. We’ve written about this extensively at the Savage Collective. This economic transformation brought with it rising economic polarization, as the returns on education and cognitive skill soared. High-skill, high-education jobs—especially in tech, finance, and healthcare—have grown in both number and compensation.

I expect something similar to happen as part of this new digital revolution—not just in work, but in leisure. We’re on the cusp of what might be called leisure polarization.

Cowen understands that AI will lead to some sort of polarization saying this:

‘Being agentic’ is an important human quality, and we expect the distribution of this trait, as expressed through actual behavior, to become more bimodal. That is, you either will work with the AIs to achieve all the more and become more agentic. Or you might respond by becoming passive and turning over more and more parts of your life to the models.

Those with the intellect, resources, and discipline to harness tools like ChatGPT will offload tedious mental labor and reinvest that time in art, history, music, and self-directed learning. For the rest of us? We’re left with brain rot.

Smartphones make it far too easy to consume the lowest forms of entertainment. And modernity, with all its comforts, has produced a peculiar kind of boredom—one that the endless scroll so effectively (and briefly) anesthetizes. It takes intellect, curiosity, and a certain moral fortitude to pursue the kinds of leisure that may soon be more accessible than ever—yet probably to be enjoyed by fewer and fewer.

What do we do about this? I don’t think there’s a clear answer. We might decide to turn off the AI for ethical, political, or spiritual reasons. But we shouldn’t do it out of fear that Tyler Cowen and his friends will get too smart or enjoy too much high-quality leisure. That kind of logic brings to mind Harrison Bergeron, Kurt Vonnegut’s dystopian short story where the government enforces radical equality by handicapping anyone with above-average intelligence, strength, or beauty.

But we do need to be sober minded and recognize that we never truly responded to economic polarization. And I suspect we won’t know how—or even begin to try—to respond to leisure polarization either.

There’s likely much more to say about how we might help people cultivate richer, more meaningful forms of leisure. Certainly, the American education system—with its near-total emphasis on human capital production—has done very little to prepare people for a life of leisure in any deep or humane sense. But I think this is enough for my first essay back.

I’ll just close with something Peter Maurin—the intellectual force behind the Catholic Worker movement— once said: “We need to make the kind of society where it is easier for people to be good.” I wonder how we might begin building a digital society that helps make it easier for all people to pursue the good through good leisure.