A Year Talking to Robots (but Learned Several Things About Being Me)
I did not begin this year with the intention of having long, reflective conversations with machines. I am a dubiously responsible adult, a university technology leader, a father, and a person who buys more books and violent video games than he has free time. I do not, as a rule, wake up thinking, Today seems like a good day to consult a probabilistic text generator about the future of learning and the ethical formation of undergraduates. Or about the future of me, for that matter.
And yet.
Somewhere between dire budget meetings, faculty technology workshops, and explaining—again—that “no, the AI cannot legally read your private meeting notes,” I found myself opening up to exactly that. Repeatedly. Voluntarily. Sometimes eagerly. Which, in hindsight, probably says more about the year than any productivity metric ever could. Give me the implant. Why not? Give it to me now.
The year, for the record, was about artificial intelligence and I suppose authentic intelligence and also I suppose about a little bit of achy insight into what we’ve lost and are losing and getting to know what’s on the horizon. Ostensibly. But mostly it was about judgment, responsibility, and the dawning realization that efficiency is a very persuasive liar. Or maybe liars are just very persuasive. See?
The Rise of the Thinking Machine and the Fall of Any Easy Answers
At first, everyone wanted to know what these tools could do. Could they write? Summarize? Analyze? Replace? Enhance? Destroy civilization as we know it?
These are understandable questions. They are also the wrong ones. As Dr. Airaudi once said, “we are standing on the whale and fishing for minnows.” I think he was quoting Joseph Campbell, who I think was referring to a Polynesian proverb.
Because once the novelty wore off, my real question became less what can this do and more what are we now tempted to stop doing ourselves. Or start doing. Really. Thinking, for example. Or teaching students, or ourselves, how to struggle productively, nobly. Or admitting when we don’t actually know any answers but are willing to find out together. Or admitting that not knowing might be our greatest knowing. We are standing on GPT, still failing to grasp Yeats. “…those were pearls that were his eyes.” Look it up. I dare you.
But yeah, this year was when AI stopped being a technology problem and became a universal character test. BTW, have you ever asked ChatGPT to assess your character? Don’t do it.
Detection: A Brief, Ill-Advised Love Affair
There was a phase—mercifully brief—when many people believed AI detection tool magic would save us. These tools promised certainty in a confusing world, much like astrology, but with dashboards. We believed that the technology people could provide the answers, or at least some succor.
They (oh yes) did not, as it turns out, save anything, or us.
What they (oh yes) mostly did was remind us that outsourcing judgment is still outsourcing judgment, even if the AI vendor brochure uses words like “insight” and “confidence interval.” Eventually, we all noticed that the tools were very sure about things they were not actually right about, which is the most dangerous kind of wrong. Well, almost.
So, quietly, without a ceremonial bonfire, we move on. Just silence and embers.
Instead of trying to catch students doing something, we started asking why what we were doing (assignments) made that something (uh oh) attractive in the first place. The year was rough. Instead of policing impossible outputs, I tried to redesign processes. Instead of surveillance, I attempted building out trust—with structure, clarity, and expectations, but still trust.
This was less efficient. It was also more instructive.
The Coalition of the Willing (No Matching Jackets, Sadly)
One of the better ideas this year was to stop trying to convince everyone at once and instead gather the curious, the cautious, and the quietly skeptical into a loose but actual coalition. I did not demand enthusiasm. I did not require ideological (or any) purity. I mostly asked people to show up, try things, and talk honestly about what happened.
What emerged was not consensus but a fragile new literacy.
People learned where AI helped and where it flattened thinking. Where it saved time and where it tempted shortcuts. Where it opened creative neural pathways and where it quietly replaced the hard work that learning actually requires. Like reading The Wasteland.
This is, incidentally, how universities are supposed to work: not by issuing decrees, but by cultivating discernment. It turns out discernment does not scale well in spreadsheets, but it ages beautifully. Like me. Relatively.
Vendors, Velocity, and Admiring a Courage to Apply the Brakes
The market, meanwhile, remains enthusiastic. Very enthusiastic. Every tool promises transformation. Every demo suggests urgency. Every sales call implies that if we did not act immediately, our institution would be left behind, presumably wandering the academic wilderness with only chalk and overhead projectors.
What this year taught me is that speed is not a value. It is a condition. And sometimes it is a condition best treated with rest.
Some of the most important leadership moments involved saying, “Not yet,” “Show me the data stewardship plan,” or “Let’s talk about FERPA before we talk about features.” These are not crowd-pleasers. They do not appear on our t-shirts. But they protect something fragile and essential: institutional trust. Or trust at all. I’ve had to learn the hard way to admire courage that applies breaks.
In an era obsessed with acceleration, restraint is a strangely radical act.
Fatherhood, Faith, and the Unreasonable Effectiveness of Paying Attention
That’s where the year got personal, whether I intended it to or not.
Being a father changes how you see systems. Children are very good at learning exactly what you reinforce, whether or not you meant to teach it. They notice inconsistencies. They absorb tone. They are unimpressed by your intentions.
This turned out to be excellent preparation for thinking about AI in education.
Because students, like children, learn what we reward. If we reward speed, they optimize for speed. If we reward output, they produce output. If we reward reflection, revision, and responsibility, they slowly, imperfectly, become reflective, revising, responsible humans.
Faith adds another layer—not as a set of answers, but as a refusal to believe that efficiency is the highest good. It insists that people are more than their productivity, that wisdom cannot be automated, and that just because something works does not mean it is working toward a right end.
These convictions did not simplify this year. They complicated it in the best possible way. I do not know if I agree with myself.
Talking to the Machines, Listening to Myself
Somewhere along the way, the machines became less oracles and more sparring partners. I used them to draft, to critique, to argue, to test half-formed ideas. Sometimes they impressed me. Sometimes they were spectacularly wrong. Kind of like me. Kind of like all of us.
What surprised me most was not what the machines reveal, but what they reflect back: a persistent concern for judgment, a suspicion of easy fixes, a desire to lead without coercion, and a deep reluctance to trade trust for control.
Apparently, I have been remarkably consistent, even when I thought I was improvising.
My Heroic Conclusion (Which Is, Of Course, Inconclusive)
This year did not resolve anything. It clarified some commitments, surfaced some limits, and reinforced a belief I hold more strongly now than when the year began: the future of learning (and me) will not be determined by what AI tools can do, but by what we decide still requires human judgment. Even if that judgement is flawed, it is flavorful.
AI did not make my work easier. It made it more honest. It forced better questions, slower decisions, and clearer values. It reminded me that leadership is not about predicting the future, but about orienting people toward it without losing ourselves along the way.
I did not plan to spend the year talking to machines.
But I did plan—whether I knew it or not—to spend it thinking carefully about what it means to remain human in the presence of powerful tools.
Which, as it turns out, is a very old question wearing very new clothes.














