Following a recent article I wrote on sport, experiential learning, and AI, I found myself pulled into a much deeper conversation.
The original argument was relatively straightforward: as AI increasingly automates the production of polished written work, experiential forms of learning – sport, simulations, leadership exercises, live projects – may become more important because they develop forms of capability that are difficult to outsource.
But the discussion that followed led somewhere more interesting.
Developmental Friction vs Obsolete Friction
A friend shared an article about creative writing education in the age of AI and made an observation that stayed with me. He argued that much of the learning in writing historically came through the friction of revision itself: identifying weaknesses, restructuring ideas, clarifying thought, refining expression.
That idea struck me because I use AI heavily in my own work.
Including writing.
If I have a meaningful insight or framework, I can now externalize it into a polished article in minutes rather than hours. The efficiency gain is extraordinary, and I have no desire to return to a purely manual workflow simply for the sake of purity.
But at the same time, I increasingly wonder whether we are entering an era where we need to distinguish between two very different kinds of friction:
- developmental friction
- obsolete friction
Some forms of friction are simply inefficient legacy processes.
But others may actually be formative.
That distinction feels increasingly important.
The Straw Bridge
Corporate team-building exercises suddenly look different through this lens.
Imagine a group of executives being asked to build a bridge from straws, construct a raft from limited materials, or solve an escape-room style challenge under time pressure.
From a purely technical perspective, the task itself is absurd. Nobody is confusing a straw bridge with real engineering capability.
But that was never the point.
The exercise matters because of what emerges under constrained conditions:
- communication
- leadership
- frustration tolerance
- adaptability
- group dynamics
- decision-making under pressure
- emotional regulation
The artifact itself is almost irrelevant.
What matters is the human being developing through the process.
And perhaps manual writing increasingly starts to resemble this category of activity.
Not economically obsolete in every context, but increasingly developmental rather than industrial.
Writing My Father’s Eulogy
A few years ago, I spent a long time writing my father’s eulogy after he passed away. A friend who had gone through a similar experience encouraged me to invest real time in it, and I’m grateful I did.
Writing it was emotionally difficult. I revised it endlessly. I struggled to articulate certain things properly. Some sentences felt inadequate. Others felt too formal, too sentimental, or simply untrue.
But looking back, I realize the writing process itself was part of the grieving process.
The friction was not accidental.
The struggle to find language was itself a form of emotional processing and meaning-making.
Today, of course, I would almost certainly use AI in that process. It would be incredibly difficult not to. The assistance is simply too powerful.
But I also suspect the experience would feel different.
The final eulogy might still be deeply personal. I would still reject phrases that felt false and iterate toward something emotionally honest. But the process of authorship would become more curatorial and dialogic rather than fully solitary.
That raises an uncomfortable but important question:
What human experiences lose something essential when friction disappears?
The Incarnational Process
A priest friend of mine recently told me that he uses AI for idea generation and research when preparing sermons, but not for writing the sermon itself.
The word he used, I think, was “incarnational.”
That fascinated me.
Because sermon writing, in that view, is not merely content production. The process itself is spiritually formative. The sermon must pass through the preacher as a human being.
And perhaps that idea extends beyond religion.
There are certain activities where the struggle itself is partly the point:
- writing a eulogy
- preparing a sermon
- journaling during grief
- writing a love letter
- certain forms of creative work
- perhaps even some forms of education itself
Not because inefficiency is morally superior, but because the process changes the person undergoing it.
Adaptive Experiences
This made me think differently about education more broadly.
For years, educational systems have largely optimized around the production of polished outputs:
- essays
- reports
- exams
- presentations
- analyses
But AI increasingly compresses the effort required to produce those artifacts.
So perhaps the educational question is no longer: “How do we produce better outputs?”
But: “What kinds of friction are actually developmental?”
Sport suddenly looks different through this lens.
So do simulations, debates, live projects, laboratories, leadership exercises, and other forms of experiential learning.
These environments introduce forms of pressure, uncertainty, iteration, failure, and adaptation that are difficult to shortcut.
They are adaptive experiences, not merely output-generation exercises.
Wisdom vs Information
The same may even apply to therapy.
Someone can study psychology academically for years and still possess relatively little depth of human understanding. Another person may have developed profound psychological insight through relationships, suffering, responsibility, failure, grief, or decades of lived experience.
That doesn’t mean formal education is unimportant. But it does suggest that some forms of wisdom are metabolized through life itself rather than purely transferred propositionally.
And this may ultimately be why AI feels simultaneously astonishing and insufficient.
It can generate language, analysis, structure, and simulation with extraordinary fluency.
But many of the capabilities we value most highly in human beings still seem deeply connected to embodiment and lived experience:
- judgment
- discernment
- resilience
- emotional regulation
- leadership
- meaning-making
- relational presence
In other words, things formed through encounter with reality rather than simply the manipulation of information.
The Educational Question of the AI Era
Perhaps that is the deeper educational challenge emerging in the AI era:
Not how to preserve old forms of friction nostalgically, but how to distinguish between friction that is merely inefficient and friction that actually forms human beings.


