I’ve been quietly noticing a trend lately. One off workshops (including our own) garner scores, if not hundreds of sign ups. Registrants attend (or don’t) and learn a thing or two and call it a day. I get it; it’s human nature to want a quick win or aha moment and call it ‘done’, but the very nature of how AI works and is evolving makes that approach not a very useful one for building real AI fluency.
It was tempting for me too. I started using AI and was ready to hang it up after a couple tutorials. “I get it!” I said to myself. Now I can go back to my life. I MAYBE would build a quick app or two, and then go back to the old way.
The shift: I stopped treating AI as something to explore and started treating it like a colleague I had to work with every single day. Once it became a daily practice, the career stuff followed. Not because I learned more about AI. Because I actually used it.
Why “keeping up with AI” isn’t the same as using it
Here’s something I say all the time in coaching: AI adoption is not a knowledge problem. It’s a behavior and psychology problem.
You can read every AI newsletter and still not be able to use these tools on a Tuesday morning when you’re staring at a deliverable you need done in an hour. That’s because understanding how AI works and being fluent with AI are completely different skills. One lives in your head. The other only comes from doing.
The people pulling ahead right now don’t have more information. They have more reps.
I came to this from a career coaching background, no technical degree, no engineering team to lean on. And what I’ve consistently found, working with professionals through this transition, is that the ones who actually build fluency are the ones who start before they feel ready. The ones who wait until they understand enough are still waiting. Waiting for enough understanding is a loop you can’t get out of from inside the loop.
What daily practice actually builds
When you use Claude or ChatGPT on real work every day, you build three things that occasional users don’t have.
Speed. Not because the tools are faster, but because you stop second-guessing every step. You know how to set up a conversation, what context to give upfront, when to push back on a bad output versus when to just take the draft and run with it. That fluency compounds. Someone using AI daily for three months moves at a totally different pace than someone using it once a week.
A track record. Daily practice means you have something to show for it. Real outputs you’ve shaped, workflows you’ve built, actual time saved on specific tasks. When AI comes up in a meeting (and it will), you have a real answer with real examples. That’s different from a vague sense that these tools “seem useful.”
Visibility. People who use AI consistently tend to mention it naturally. They share what worked. They help a colleague when the tool produces something confusing. That organic, low-key visibility is how most people become the person their team turns to when AI questions come up. Which is exactly where you want to be right now.
The compounding effect: you’re building judgment, not just skill
Here’s what most people miss about daily practice: you’re not just getting better at AI. You’re developing judgment.
You start to know which tasks AI handles well and which ones need you fully in the loop. You develop a feel for when a first draft is 80% there versus when it’s missed the point entirely. You get faster at giving feedback and better at knowing what “good” looks like for your specific work.
This judgment doesn’t transfer from reading about AI. You can’t get it from YouTube. It only comes from the accumulated experience of doing it enough times that you know what good looks like for your actual job. That’s the expert collaborator role. Research on knowledge workers and AI consistently finds that AI amplifies capability but can’t substitute for domain expertise. Daily practitioners build the collaboration muscle. Occasional users stay in the experimental phase indefinitely.
What 20 minutes a day actually produces
Here’s your permission slip: you do not need a big dedicated AI block on your calendar. You do not need to work nights and weekends on this stuff. You need a small repeatable habit attached to work you’re already doing, during your actual work day.
Twenty minutes a day, five days a week, on real tasks. Here’s what that actually produces over time.
In the first two weeks, you collect wins. Not big ones. A draft that saved you thirty minutes. A summary that caught something you would have missed. One workflow that now takes half as long. These wins matter because they’re evidence. And evidence is what converts occasional interest into an actual habit.
By weeks three through six, you start developing intuition. You know how to give Claude enough context that the first output is closer to what you need. You’ve stopped treating it like a magic box and started treating it the way you’d work with a capable new hire: give context, set expectations, evaluate, iterate. (This is literally coaching. Which, as a coach, I find delightful.)
By month two and beyond, the compounding shows up. You have specific examples you can draw on when AI comes up at work. You have opinions about which tool works best for which tasks. You’ve built something, solved something, or saved something you can actually point to. That’s the career capital daily practice produces.
The move I find most powerful, and I tell this to everyone: anything you do more than once, write down how you approach it, then teach Claude your method. The next time, you start from there. You’re not just using AI. You’re building a library of repeatable workflows that gets better every time you run it.
The window that’s open right now
I want to be direct about the career piece here, because I think it matters and it tends to get either oversold (fear-based) or undersold (vague).
Most organizations are in the middle of figuring out AI. They know they should be doing something. They’re not sure what. The people who are going to lead that conversation at the team level, the department level, the whole org, are not going to be hired from outside. They’re going to be the people who are already there, already doing the work, and already visible as someone who has a clue.
I’ve watched this happen with people in our community. Someone starts using Claude every day on their actual work. They mention it in a meeting. They help a colleague get unstuck. Three months later they’re running the AI pilot for their department. They didn’t apply for anything. The opportunity found them because they were ready.
That’s the window. It won’t stay open forever. As consistent AI use becomes table stakes (and it will), being someone who uses these tools daily won’t be a differentiator. Right now, it still is.
The professionals building their AI fluency through daily practice today are positioning themselves for conversations that are going to happen in the next six to twelve months. Not by claiming expertise they don’t have, but by actually having it when the moment comes.
How to start (or restart, if you stalled)
Pick one task you do every day or every week. Something with a clear output. Something you’ve done enough times that you’d recognize immediately if the AI version was good or bad.
Take that task to Claude tomorrow. Give it real context: who you are, what you need, who will read it, what a good result looks like. See what comes back. Edit it. Notice what worked and what didn’t. Don’t close the tab when the first output isn’t perfect. That’s actually the moment to stay in it.
Do that again the next day. Same kind of task, not necessarily the same one. Build from there.
The goal in week one is not to transform your workflow. It’s to create one piece of evidence that this is worth coming back to. One good result on real work is worth more than a month of reading about what AI can do.
If you want to see what this kind of daily practice can lead to career-wise, How to Become the AI Person on Your Team walks through exactly how that positioning happens.
The people figuring this out together
The hardest part of daily AI practice usually isn’t the tool. It’s doing it without a reference point. You try something, get output that’s okay-ish, and you really don’t know if you’re on track or missing something obvious. That uncertainty is what quietly makes people stop. Not because AI failed them. Because the isolation made the friction feel like personal failure.
The professionals I see making the fastest progress aren’t doing it alone. They’re in community where they can share what’s working, borrow approaches that wouldn’t have occurred to them solo, and get the reminder they need that the friction they’re feeling is normal.
If you want to practice alongside people who are doing this every day and being honest about what they’re learning, that’s exactly what happens in our community. Join us for 2 weeks to check us out if you’d like.
The habit goes faster when you have somewhere to bring the questions.