The Conversation That Continues

Friendship as a model for AI

I have an old friend. We’ve known each other for decades, and something surprising has never happened: we’ve never arrived at complete understanding.

There was no moment where we could say, “Well, that’s it. We’ve figured each other out. No more questions.” Instead, our understanding deepens in motion—through updates, disagreements, shared memories, changed minds, and the simple act of continuing to talk.

Lately, I’ve started to think about AI the same way.

Not as a tool you master once and then put on a shelf, like learning a software shortcut or fixing a leaky faucet. But as a conversation—ongoing, evolving, and never quite finished.

Why “mastering AI” can feel strangely unsatisfying

There’s real skill involved in working with AI well. You learn how to prompt clearly. You learn how to evaluate what you’re being told. You build judgment about what to trust, what to verify, and what to discard.

But here’s the catch: AI is changing. Constantly.

The models update. The interfaces shift. New features appear. Your work changes, too. Your life changes. The questions you ask today won’t be the questions you ask next year—maybe not even next month.

So the old idea of “I’ll learn this, then I’ll be done” starts to feel less like ambition and more like a misunderstanding of what the relationship actually is.

The goal isn’t to reach a final state of mastery. The goal is to keep the conversation useful.

Naomi’s plateau that never came

Naomi is a landscape designer who has used AI for three years. When she started, she assumed she’d climb a learning curve, hit a plateau, and settle into competence.

“I had this fantasy of reaching a plateau,” she told me. “Three years later, I’m still learning.”

And that’s not a failure. That’s simply what happens when you’re engaging with something that is both responsive and evolving—especially when your own work is creative, context-heavy, and full of constraints.

What changed for Naomi wasn’t the technology. It was the way she measured success.

She stopped tracking progress by “how much I’ve mastered” and started tracking it by “how useful was the conversation today?”

Some days, AI helps her see what she might miss—an overlooked design option, a clearer way to explain a plan to a client, a faster path through a pile of research.

Other days, it leads her astray—offering confident nonsense, missing key context, or pushing toward generic solutions that don’t fit the specific land, budget, climate, or human beings involved.

Both are part of the conversation.

The wabi-sabi approach to working with AI

In Japanese aesthetics, wabi-sabi finds beauty in imperfection and incompleteness.

A tea cup with uneven glaze. A weathered wooden beam. A garden that is never quite “done” because seasons keep changing it. The point isn’t flawlessness—it’s aliveness.

Our relationship with AI can be wabi-sabi, too.

It will always be partial. Sometimes brilliant, sometimes clumsy. Often helpful, occasionally wrong in ways that are almost comical. And while that can be frustrating, it can also be grounding—because it reminds us we are not dealing with an all-knowing oracle.

We’re in a relationship with a tool that behaves like a collaborator: offering possibilities, requiring discernment, and improving through iteration.

When you adopt this mindset, you stop demanding perfection from the exchange and start paying attention to what the exchange makes possible.

Skills for an ongoing conversation (that don’t go obsolete)

If AI is less like a course you complete and more like a relationship you develop, then the most important “AI skills” aren’t just technical. They’re human.

A few qualities make the difference:

Humility.
Not the performative kind, but the practical kind: “I might be wrong.” “This might be incomplete.” “I should verify.” Humility keeps you from outsourcing your judgment.

Curiosity.
Curiosity keeps the conversation alive. Instead of using AI only to confirm what you already believe, you ask: “What else could this mean?” “What am I not considering?” “Give me three alternatives.”

Patience.
Good results often come after a few turns. The first answer may be sloppy. The second might get closer. The third might finally fit. Patience is what turns AI from a vending machine into a thought partner.

Flexibility.
Today’s best workflow might be obsolete in six months. Flexibility helps you adapt without feeling like you’re starting over every time the tech shifts.

These aren’t just AI skills. They’re life skills. That’s part of why this conversation matters.

The real invitation inside the uncertainty

My old friend and I will keep talking. New experiences will change the way we interpret old stories. New challenges will reveal blind spots. That’s what friendship is: not a destination reached, but a journey shared.

AI will keep changing, too. The conversation has no final chapter.

And perhaps that’s not a burden. Perhaps it’s an invitation—to stay engaged, keep learning, and remain in relationship with something larger than ourselves: creativity, knowledge, curiosity, and the messy, beautiful process of becoming.

The conversation continues. And so do we.