The Robots Have Won…Sort Of.

Dear colleague,

I was recently disturbed by a new AI tool that popped up in two of the handful of news sources I read on a regular basis. Basically, you give the tool a link to an article, a PDF, or a Google document, and with the click of a button, the tool creates a “deep dive” podcast in which two eerily human-sounding “personalities” discuss the contents of the source material.

Let me show you what it sounds like (see 1:38 through 8:48 in ​the video below​) when I feed it ​my article on Mini-Sermons​.

I don't have polished thoughts on all of this today. Instead, I'll share my rough-draft reactions to this NotebookLM breakthrough and the broader trend of rapidly improving AI software.

First up, there's this: I do feel threatened by AI software like this.

I think up to this point, I've tried putting on a brave face with AI, as both a teacher and a writer. But as generative technologies have rapidly infiltrated the Internet, its software services, and its connected devices, while at the same time (and as a result of this rapid spread) become remarkably sophisticated, I've been unnerved.

Do I continue to have faith in the undying importance of human-focused classrooms, human-facilitated learning, human-created writing? I do.

But do I struggle in this faith at times, when encountering new marvels like the “podcast” above? Absolutely.

I mean, here's the blunt feedback I have on the NotebookLM “podcast” meant to “deep dive” into my blog article: Despite a few glitchy moments and awkward transitions, the quality of interaction between the podcast “hosts” is more interesting and engaging and quick-witted than I am.

If I had been interviewed by someone about mini-sermons (strategy #4 in ​The Will to Learn​), I couldn't have done it as well as these folks do. If two teachers who love my work had set out to create a high-quality discussion of my article on mini-sermons, I don't know if they could have come up with something as comprehensive or engaging as this. In other words, the quality, though not perfect, is excellent.

And if the rapid adoption and improvement of these technologies has been any indicator, that excellence will continue to improve.

Rapidly.

So that disturbs me. In a world of eyes and ears trained by algorithms to seek maximally engaging content, it's unsettling that robots are more engaging than me. (And when I say “more engaging than me,” I'm speaking technically, e.g., tone of voice, balancing humor and insight, pacing variation between lingering and moving on from ideas.)

Furthermore, these robots mostly do “get” what I was after — better than most humans probably do when they first encounter my work. They've taken my imperfect writing and developed an “understanding” of it that's pretty close to spot on, and they've done so in less than 10 minutes.

So as a teacher and a writer, I feel a bit outmatched by this technology. It's like that feeling you get when a student comes into your room who is more gifted or intelligent than you. Except that it's not like that at all because this “student” is an always-consuming, always-improving software program that has no soul and does not sleep and is owned by a for-profit corporation.

As I write that, I laugh. But maybe that's just to keep from crying.

On the other hand, nothing actually happened when this AI podcast was created because the product is not the point.

On the flip side, nothing of substance happened when this “podcast” was created from my blog article.

• No human being learned anything, no human soul changed, no new meaning entered the world, no real-world relationships were formed.

Nothing actually happened that was lastingly important.

And here I think we get to what AI is probably going to force us teachers and our educational systems to come to terms with: the products have never been the point.

When I teach students a unit, the students' mastery of that unit is not the lasting worth of what we've done together. The lasting worth is that, as we worked toward mastery together in my classroom, we were transformed.

• Our souls were altered, in that:
  • We formed new relationships — with ideas, with each other. The social facets of our souls.
  • Our minds were broadened, deepened, sharpened. The intellectual facets of our souls.
  • We experienced a range of emotions. The soul is emotional.
  • Our beliefs about learning were affected, and through these beliefs our wills.
  • Our physical bodies changed, in that our brains formed new neural pathways. The soul includes the body.

This type of soul transformation — which is the normal stuff of education, the stuff that happens day to day, whether we intend it or not — can't happen through feeding an article into NotebookLM. Soul transformation can't be done to human beings; human souls just do transform as they engage in the work of learning.

But I think an error we've made, for a long time now, is to mistake the products of learning (the test results, the finished essays, the completed projects) with learning's fruits.

So could I double or triple or quintuple my writing output on this blog by using tools like ChatGPT? Could I start a podcast for this blog using tools like NotebookLM?

That all depends on what we consider “output.”

• If output is just the words that get published, the content cranked out, then sure — I could.
• But if output isn't that — if it's actually the transformation that takes place in me through the writing process and the transformation that takes place in you as you seek to understand my words — then no, I could not.

My Commitment to You

Here in DSJR-land, I'm going to keep making stuff for you, colleague. I'm meaning, me — I'm going to do it.

And what I hope is that these human-crafted words continue to provide a value for your soul that software-generated content cannot.

I'll also keep stoking the fire of my faith in what an education fundamentally is: the promotion of the long-term flourishing of a learner by teaching that learner toward mastery of the disciplines. It was an old-fashioned idea when I first articulated it, and it's perhaps more old-fashioned than ever in a time where machines are gaining “mastery” at rates humans simply cannot.

Old-fashioned, yes. But able to stand the test of time — even times like these? Yes again.

Teaching right beside you,

DSJR