The Wave Is Moving
A month ago I wrote about 2026 being the year the conversation around AI finally changed. I talked about Frontier Firms, redesigning work, all of it still true. But something happened since then that I didn’t expect: my own workflow changed faster than I thought it would.
Since Valentine’s weekend, I’ve been in overdrive. Not the unsustainable, forgot-to-eat-lunch kind. (Okay, sometimes that kind.) The kind where something clicks and suddenly the friction you’d been working around just isn’t there anymore. The tools caught up to the ambition, and now I’m the bottleneck — which, honestly, is a much more interesting problem to have.
Claude Opus 4.6 is the inflection point for me. I’ve used every major model over the past two years — some for months at a time, some for exactly one frustrating afternoon — and this is the first one where I stopped thinking about the tool and started thinking only about the work. The difference isn’t benchmarks or context windows. It’s that the gap between what I’m trying to do and what the model can hold in its head effectively disappeared. I’m not prompting anymore. I’m collaborating. My wife and I planned our Saturday night around it — not Netflix, not dinner out, just the two of us hanging out with Claude, building things.
Pair that with Claude Code and OpenClaw, and the whole shape of my day is different. I went from the chatbot-and-CLI era of 2025 — where AI was a better autocomplete that occasionally surprised you — to something that feels more like a creative partner with its own working memory. The workflows aren’t just faster. They’re structurally different. I’m defining outcomes and reviewing results instead of writing every intermediate step. It’s like going from hand-drawing animation frames to directing the movie.
This is changing how my team operates too. The conversations are shifting from “how do we build this” to “what should we build” — which is where the interesting decisions live. When execution speed stops being the bottleneck, strategy becomes the actual job. That’s a bigger shift than it sounds like. It’s also way more fun.
I want to be honest about where I sit on the curve. I’m not building foundation models. I’m not doing cutting-edge ML research. There are people who are already living in a different world — the ones deep inside labs and frontier research who don’t post about it because they don’t need to. I’m not in that world. That’s fine. What I can see from where I stand is that the trajectory is steep and the direction is clear. We are moving toward a layer of abstraction where the primary skill is defining what you want — clearly, precisely, with taste — and the execution layer handles more and more of the how. That’s not science fiction. That’s what’s happening right now, in real workflows, for people willing to rebuild their process around it.
The other thing that changed: I finally built a content pipeline that works. For years I’ve had things to say and no system for saying them consistently. I’d jot ideas in notebooks, dump thoughts into text files, and never take the time to compose and share any of it. The ideas were there. The pipeline wasn’t. That’s fixed now. You’ll be hearing from me more — here, on LinkedIn, wherever the conversation is happening. Not because I think I have all the answers, but because thinking out loud is how I’ve always learned best, and the pace of change right now demands it.
There’s more coming. A lot more. Consider this the “hey, I’m here” post before the actual wave hits.
See you soon.