Your Relationship with AI: Don’t be an expert. Don’t (ever) be a tool. Be a Player.
~ 900 words. 4 minute read.
Have you noticed this pattern: Someone discovers AI - truly discovers it, interacts with it - and one of two things happens.
They become overwhelmed and step back. Or they become confident and start performing.
According to Faisal Hoque’s thesis in ‘Transcend - Unlocking Humanity in the Age of AI’, neither one of these is the relationship worth having.
I believe that the people best positioned to get the most out of AI are not who we expect them to be.
The Expert Problem
In a previous post, I wrote about the Dunning-Kruger effect - that delightful cognitive trap where low competence produces high confidence, and real expertise tends to produce humility. There’s a version of this playing out in how people engage with AI right now, and it’s subtle.
We tend to approach a tool from a position of competence. We have a job title, a framework, a set of established heuristics. We know what ‘good’ looks like. And when we sit down with AI, we immediately filter everything through that lens - prompting for what we already expect, validating what we already believe, using a powerful new instrument like a slightly fancier search bar.
This isn’t stupidity. It’s very human. But it’s a waste.
Shunryu Suzuki, the Zen Buddhist teacher who brought mindfulness practice to the West, had a phrase for the antidote: Shoshin. The Beginner’s Mind.
“In the beginner’s mind there are many possibilities. In the expert’s mind there are few.”
Many possibilities.
The Beginner’s Mind isn’t about pretending that you don’t know things. It’s about suspending the need to already know, so that something genuinely new can enter. It’s an active posture. A choice.
The expert approaches AI with answers already half-formed. The beginner arrives at the conversation with a question and stays curious long enough to be surprised.
Being surprised, it turns out, is the whole game.
The Tool Problem
On the other end of the spectrum, there’s the transitional crowd. They use AI instrumentally and efficiently. Input in, output out. Summarize this. Rewrite that. Generate five options. Pick one. Move on.
Nothing wrong with this. It has clear value. But it’s also not a relationship. It’s more like vending machine behavior.
Dutch historian Johan Huizinga spent his career arguing that play is not something we do in addition to serious human activity. It is serious human activity. His concept of Homo Ludens, the Playing Human, suggests that human culture emerges from play. From games. From the willingness to enter a bounded, exploratory space and see what happens, without needing the outcome to be predetermined.
This powerful concept seeks to unlock human creativity, but as much as we crave innovation, our corporate life with its quarterly forecasts, goals and objectives has squashed every last ounce of it. We simply don’t have time for games - especially if the outcome is uncertain.
Our chance has come to re-discover our Homo Ludens, and it lies in our relationship with AI. Next time you write a prompt, approach it with the mindset of “let’s try something” instead of ‘give me X”.
Permission to play.
Treat a conversation with AI the same way a child treats a new toy - not reading the manual, just picking it up and seeing what it does (in product design, we might call that ‘user testing’, but it’s allowing the user to be Homo Ludens. This is about allowing the creator to be Homo Ludens.)
The people I’ve seen get genuinely transformative value out of AI are the ones who play. Who iterate. Who follow a thread somewhere unexpected and don’t close the tab out of impatience. Playfulness is not the opposite of productivity here (as any corporate setting of what quickly sounds like an antiquated way of doing things might prescribe). It’s a prerequisite for it.
The Wildcard
Now here’s the part I find most fascinating (and most personal) when exploring what ‘relationship with AI’ could mean.
In these early stages, we talk a lot about having to learn how to use AI. It’s new tool and we have always approached new things with a learning mindset - it’s what we humans are proud of! But in this case, the very essence of this new tool is premeditated on learning us.
Large language models are probability engines. They predict. They pattern-match. They complete. Given a prompt, they calculate the most likely continuation based on everything they’ve ever been trained on. This makes them extraordinarily good at the expected, the conversational, the well-trodden.
And here’s what this means: the people who will most consistently surprise an LLM are the people who don’t think in the ways anticipated by the training data.
Breakthroughs happen when we surprise AI.
When he was three years old, my son was diagnosed ASD or Autism Spectrum Disorder. He processes the world differently - sometimes in ways that look like obstacles to the people around him, and sometimes in ways that stop me cold with their precision or their ‘lateral beauty’. He recognizes and applies a pattern where we would not have. He asks questions that don’t occur to others. He resists - or doesn’t perceive of - the obvious framing.
My logical conclusion then is that neurodiverse people are going to be among the most important contributors to expanding our relationship with AI and to achieving real breakthroughs. Precisely because of the fact that they, yes, literally, think differently.
When everyone else is getting a very polished, very predicted response, the person asking an unexpected question from an unexpected angle is discovering something new. They are, in the truest sense, jailbreaking the probable.
We spend a lot of time talking about AI’s potential. We spend less time thinking about who will unlock its potential first, and whose potential it will unlock first.
By the way, here’s an exquisite moment of observing the universe tickling itself - my son’s initials (not planned with any of this in mind at the time of his birth): A.I.
What this asks of us
What’s the true relationship potential we have with AI then?
It lies in the Beginner’s Mind: Accept, nay, seek many possibilities.
It’s unlocked by Homo Ludens: Curiosity and playfulness over efficiency.
Start watching the kids in your life, who, as a stroke of evolutionary genius are gifted the Beginner’s Mind and playful experimentation as their highest evolved attributes right off the bat. Maybe watch especially the ones who’ve been told by the world that they are ‘different’ - they may be better prepared for humanity’s future than you are.
Where on the Dunning-Kruger curve do you fall when it comes to AI? The best answer on Dunning-Kruger is always(!): Lower than you think.
And that’s good news.
