The More I Understand AI, The More I Want to Do

There’s a specific kind of frustration that only kicks in when you start getting good at something. Not the early frustration, where everything’s confusing and you don’t know where to start. The later kind. Where you can suddenly see ten things you want to build and you only have time for one.

That’s where I’ve been this week.

Anthropic’s Academy and the Prompt That Changed How I Think

I’ve been working through Anthropic’s Academy this week. Claude’s own training ground, basically. And if you’ve been treating AI like a smarter search engine, this is the thing that’ll reset how you think about it.

The big shift for me wasn’t any single lesson. It was realising that the gap between what AI can do and what most people use it for is enormous. Not because the tools are hard. Because most people don’t know what to ask for. You can’t prompt for something you can’t imagine yet.

That’s the whole game. Imagination, backed by education.

Vibe Coding: Building Things I Couldn’t Have Built Last Year

Alongside the Academy, I’ve been vibe coding. If you haven’t heard the term, it basically means describing what you want to build in plain language and letting AI generate the code, then iterating from there.

I’m not a developer. I’ve never pretended to be. But this week I built things that would have required me to hire someone twelve months ago. That’s not a small thing.

It’s not magic. You still need to understand what you’re building well enough to spot when it goes wrong. But the barrier is lower than it’s ever been, and it keeps dropping.

n8n: Automations That Actually Run While I Sleep

The other thread running through this week has been n8n. I’ve been building more workflows and, more importantly, building the monitoring around them. A launchd service on the MacBook polling every five minutes, keeping an eye on things so I don’t have to.

The social media side is expanding too. Bluesky automation is live. More platforms in the pipeline. Content that used to take manual effort now moves on its own.

None of this happened in a week. It’s been building for months. But this week felt like a tipping point, where the foundations started paying out.

The Weakest Link Is Always the Same

I nearly posted something on LinkedIn this week about all of this. Bottled it at the last minute and wrote this blog post instead, which feels very on brand.

But the LinkedIn draft had a point worth making here. The thing holding most people back from getting the most out of AI isn’t the technology. It’s not the cost. It’s not the learning curve, not really.

It’s imagination. And imagination is downstream of education.

The more you learn, the more you can picture. The more you can picture, the more you want to build. And the more you want to build, the faster the week goes when you’re actually doing it.

That’s where I am right now. Ten things on the list, time for one.

Which is, honestly, a good problem to have.