Do me a favor. Close your eyes.

That's how I open Blink, the story I told last week at the Santa Fe Story Lab, an event supported by the Friends of the Santa Fe Public Library's ongoing experiment in getting people to stand in a room together and tell each other true things. About 150 folks showed up to hear 6 of us storytellers, with stories about family, work, loss and discovery. The host, Sarah Alpers, has created a wonderfully supportive context for community story sharing.

The piece leans on the one thing I remember from Walter Murch's In the Blink of an Eye: that when your eyes move from one point of attention to another, you blink, and your brain uses that blink the way a film editor uses a cut. You don't notice thejump. You just notice you're somewhere new.

I like that idea because it gives you, standing on a stage with no notes, something like an editing rhythm — you can find your own cut points, your own blinks, in real time, in front of people who are blinking along with you. [Fill in: the actual turn in the story, what the Murch bit was doing narratively, how it landed.]

Here's what doesn't happen, though, no matter how good the metaphor is: nobody can undo a blink. There's no take two. That room, that Tuesday, that particular configuration of strangers and friends who'd wandered in — that's the whole show. It happens once, in a body, in a specific place, at a specific hour, and then it's gone except for whatever the people in the room decide to keep.

I've been thinking about that unrepeatability all week, because I've also been reading Cory Doctorow's new book, The Reverse Centaur's Guide to Life After AI, which is a different kind of animal but rhymes with this in a way I didn't expect. More on that in a minute. First, the orality thing, because I can't help myself.

We are, apparently, going backwards. On purpose. Sort of.

Walter Ong spent a career arguing that the shift from oral culture to literate culture wasn't just a new delivery mechanism for the same old information — it rewired how people thought. Oral cultures held knowledge in bodies, rhythms, formulas, and relationships, because that's the only storage medium available when nothing gets written down. Literacy let you set an idea down, walk away, and come back to the identical idea later. It's a completely different relationship to memory and truth.

There's a piece that ran in The Atlantic this month (here's a copy if you don't have the subscription) making an argument I think is relevant: that we're now living through a return swing of that same pendulum, a drift back toward orality, driven this time not by the absence of writing but by machines that read, summarize, and speak text back at us so we don't have to sit with it ourselves. The argument isn't that literacy is disappearing exactly — it's that our habits of mind are sliding back toward something more oral: quick, social, performative, keyed to a speaker's voice and cadence rather than to a fixed page you can quietly interrogate.

If that's even half right, it's a strange loop to be standing in. Ong wrote about the passage from orality to literacy as a hinge point in human history. We may be living through a second hinge, except this time the "orality" on offer is manufactured — a chatbot's smoothed-out cadence, a synthetic voice reading you the news, a summary performed at you rather than a story told with you. That's not orality. That's a puppet show wearing orality's clothes.

Which is exactly why what happened at the Santa Fe Story Lab matters more than it used to, not less. Real oral storytelling — off-script, in a room, with an audience that can see you sweat and can make you change your ending mid-sentence because of how they're leaning forward or not — is not a format. It's an encounter. Nobody has figured out how to synthesize an encounter, and I'd bet against anybody figuring it out, because the whole point of an encounter is that it can't be pre-computed. As machines get better at handling our communication chores — email, memos, meeting notes, even a decent first draft of a eulogy — the ancient, embodied, unrepeatable version of story doesn't become a nostalgia act. It becomes the thing we go to precisely because it can't be faked. We may be circling back to orality, as the pendulum swings suggest, but the honest, sweat-and-blink kind is going to be worth more, not less, the more the ersatz kind floods in around it.

Now. About the centaurs.

Doctorow's book gives a name to something a lot of us have been feeling without a word for it. A "centaur," in his telling, is a person using a machine to become a better version of themselves — augmented, in charge, setting the pace. A "reverse centaur" is a person conscripted into serving the machine's pace instead: Lucy and Ethel at the chocolate belt, except now the belt is your inbox, your patient load, your grading queue, your customer-service quota, and it never, ever slows down for you. Doctorow's real target isn't the technology so much as the business model wrapped around it — the investor logic that needs you working at superhuman speed to make the valuation math pencil out.

He also addresses an industry in such a rush to "win" the lottery of trillions of dollars in investment, that it is bulldozing the planet to get there. To win, they need more compute, which means mega-data Centers, and the use of technologies that suck up so much electricity, water, carbon, rare earth minerals, toxic chemistries, that it is expediting the threat to all life on earth. It doesn't have to be this way, this is a particular process related to global hegemonies, power, and the economic gambles of the 1/100th % of leaders in industry (primarily US) or government (primarily China). I honestly believe that over a longer time span, we could evolve the more sustainable ways of making these tools available.

I think we need real regulatory teeth on this industry, and I'm not allergic to the word "nationalize" when it comes to infrastructure this consequential. The public interest and the venture-capital return schedule are not the same thing, and pretending they'll naturally align is how we got here.

And yet — I have become an AI user, mainly because I like understanding what new technologies might be useful for in the context of community and social justice, and in no small part because I need the productivity and support, I'm not up for working 12 hour days anymore, and I feel the clock ticking on my days of any kind of productivity, and I am not going to pretend otherwise. That's the dilemma for a lot of us on this side of sixty-five, and it's not a hypocrisy so much as a bind. We watched our institutions get hollowed out — pensions, public education, affordable care, a labor market that used to reward loyalty instead of churning people out at fifty-five. We know, in our bones and in our politics, that the economics underneath all this are lopsided, extractive, and often actively hostile to older people, disabled people, poor people, the planet. And at the same time, "assistance" — genuinely useful help with the parts of a task that used to eat a whole afternoon — is enormously tempting when the afternoon isn't guaranteed the way it once felt like it was.

So here's where I land, at least this week: the tool isn't necessarily the problem. The ownership structure is the problem. I want the option to be a centaur — to use a machine to get the tedious 20 percent of a task out of the way so I can spend the 80 percent that matters, the part only a human can do, on the part that matters. What I don't want, and what I think we should organize against with everything we've got, is a system engineered to make me — or anyone, but especially anyone whose margin for error is thinner than mine — into the reverse kind: conscripted, sped up, dependent by design, and billed monthly for the privilege.

None of that changes what happened in that room at the Santa Fe Story Lab. You can't outsource a blink. You can automate a summary of a story, but you can't automate the thirty seconds where a room goes quiet because somebody just said the true thing out loud, in front of people, without a script, with nowhere to hide.

That's still ours. I intend to keep doing it as long as I've got a working set of eyelids — and I'd argue the more the machines take over everything else, the more that one un-outsourceable thing is worth protecting, funding, teaching, and showing up for.