I just got back from four days in Cancun, Mexico. The trip was part of an Improving recognition I received — the Improving 100 award for technical innovation in 2025. (I wrote about what that recognition meant to me, and what thought leadership actually is, in a previous post — I won’t retread that ground here.) What I want to talk about is what I noticed on this trip and what I brought back (besides my awesome trophy and great memories!)
The Spanish Thing
One of the things I’ve been doing for years is practicing Spanish on Duolingo. A few minutes a day, most days. I took a detour into chess lessons for the last several months and set Spanish aside, but I still have all those words rattling around. Cancun was a chance to see how many of them actually stuck.
I tried to use Spanish in most of my interactions at restaurants and the hotel. And I definitely used the wrong word for something at least once. I could tell immediately: the other person’s expression shifted to polite confusion. I’d try to explain what I meant, they’d give me the right word, and I’d file it away. That’s the real test, not the app.
What made it stick wasn’t the lesson. It was the moment of getting it wrong, realizing it, and then hearing the correction in context. Public. A little awkward. Memorable.
That’s the same dynamic I’ve been navigating with AI for the past couple of years.
Thought Process Over Prompts
When I was recording those weekend videos for our internal Improving channels (showing how I was using AI tools on real problems), I wasn’t focused on sharing my prompts. Prompts felt too narrow, too dependent on the exact version of a tool at a specific moment. What I wanted to show was how I was thinking about interacting with AI.
How do you break a problem into pieces that a model can actually help with? How do you stay in the driver’s seat when the tool is very confident and sometimes confidently wrong? How do you build something across multiple sessions without losing the thread?
Those questions mattered more to me than “here’s the prompt I used.” And over time, that turned out to be a useful distinction. Because the tools kept changing. Better context windows. System prompts. Then skills that let you package up your approach and reuse it. The prompts from a year ago are less relevant today. But the thinking behind why I structured things a certain way? That still applies.
When Abstractions Become People
I knew some Improvers followed my blog or had watched those videos. I’d heard from a few of them over time. But most of my sharing happened in a kind of void: you put the content out, you see a few reactions, and you move on. You don’t really know who it’s reaching or what they’re doing with it.
Then I got to Cancun and started meeting people in person. Improvers from Argentina, Guatemala, Canada, India. Some of them had read my posts. A few of them mentioned specific things they’d tried or adapted based on something I’d shared. That’s a different experience than getting a thumbs-up on a channel post. You’re suddenly talking to a person who made something concrete out of an idea you put into the world.
The Virtuous Cycle
Here’s something about learning in public over time: it compounds.
When I look back at some of the lessons I was sharing a year ago, I can see how far the thinking has evolved. Some of what I shared then I’d say differently now, not because it was wrong, but because I’ve internalized it, refined it, found better tools or better approaches since. And I’ve kept sharing as that happened.
That’s not a problem. That’s the point. The content from a year ago wasn’t meant to be the final word. It was a snapshot of where I was and what was working at that time. Sharing it didn’t lock me in. It gave me something to build on.
And when I hear from other Improvers who’ve done the same thing (tried something I shared, modified it, found a better way, and came back to tell me), that’s the cycle working. They’re not just consuming content. They’re pushing the thinking forward.
What I’m Carrying Back
I came home from Cancun genuinely motivated. Not because of the resort or the beach, though the decompression was exactly what I needed. But because of those conversations. Because of the confirmation that the sharing matters even when you can’t see it mattering.
I want to keep finding ways to share my learning: as I go, when it’s messy and not figured out yet, and also a few months in, when things have settled, and the lesson has clarified. Both kinds of sharing are useful. The early, scrappy version shows that it’s okay not to have the answers. The later reflection shows what the journey looked like after you’ve walked it.
And now, having met some of the people who’ve been on the other side of these posts and videos, it makes things more concrete. I know a little more clearly who I’m writing for.
What would you share if you knew someone out there was building on it?
Leave a Reply