What I’ve learned from watching smart people use AI
I was on a call recently with someone who's been writing about technology for twenty years, he’s whip smart and has always been an early adopter.
He uses Claude every day. When I asked how he structures his working week around it, there was a long pause.
"I don't, really. I just open it when I need it."
He's not unusual. He's the norm.
It’s fair to day I’m all in on this stuff. AI is my working day. It's what we do - we're in it constantly, trying things, watching what breaks, sitting with clients while they experiment in real time. That means more space to ponder on this stuff than most people get. More time to notice where things go wrong. And more examples.
And the thing we keep seeing is this: there's a difference between having a tool and having a practice.
A few weeks ago I was on a call with a retail founder - sharp, commercially minded. He'd been trying to use AI to find his largest transactions for a specific product across six months of email orders.
It kept returning the three most recent. He asked again. Same result.
"It's like asking someone to do a job and they half do it and keep coming back saying they've done it."
He's not wrong. But the problem wasn't the tool. It was that nobody had told him where it breaks and why setup matters as much as the ask.
That's not a technology problem. It's fixable in an afternoon.
We're working with a founder right now who had, without realising it, been building something valuable for two years.
He'd been using AI as a personal thinking partner; working through decisions, processing difficult conversations, developing his approach to leading his team. His compensation philosophy. How he thinks about developing specific people. The patterns he keeps returning to as a leader.
All of it sitting in a chat window. Accessible only to him, in a format that was neither searchable nor deployable, and which would eventually just be gone.
We're extracting it, structuring it, turning it into something he can actually use. His accumulated thinking becoming a business asset rather than something that lives in his head and dies with his laptop.
That's the version of AI adoption I find most interesting. Not the tool that replaces a person. The one that multiplies what they already know.
Last week we sent a client the first output from a tool we've been building together. A full sector briefing and two-week campaign pack — research, framing, three LinkedIn posts in their voice, outreach copy, a CRM check on latest status.
His response: "This looks brilliant — could we start now/Monday?"
Minutes to generate. Useful because it was built to know them - their language, their positions, their existing relationships. Generic AI could have done something similar with an hour of prompting and correction. This needed none of that.
That's the difference between a tool and a system.
I was in the room in the late nineties when people were saying nobody would ever buy shoes on the internet. I was very very young obviously.
Didn't move fast enough, didn’t really undestand the upside. Watched the window and then watched it close.
I think about that a lot right now. Not because I'm certain how this plays out - I'm not. But the cost of experimenting is almost zero. The cost of waiting is compounding every month.
We work with founders and senior leaders on exactly this. Not only the strategy layer but the operational layer.
What does your working week look like with AI genuinely woven into it? What decisions could it pressure-test? What knowledge is sitting in your head - or a chat window - that your business should own?
Living in this every day means we can shortcut a lot of the trial and error for you.
Here’s a thought starter on how to start building a practice rather than just using a tool.
→ How to Build an AI Practice Into Your Working Week
If any of this resonates — I'd love to talk. Not a sales call, just a conversation to help me understand opportunities (and hurdles).
