The Tools Are
Ready. Are We?

A year ago, building the Whist Scorer would have been a weekend project — schema, state management, deployment, the works. Last month it took twenty minutes, most of which was spent dealing cards. The tools crossed a threshold when I wasn’t watching.

And yet most people I talk to are using the same tools to write slightly faster emails. The gap between what’s possible and what’s practised is the widest I’ve seen in my career. That gap isn’t technical. It’s a skill we haven’t collectively learned yet.

The bottleneck moved. It used to be “can it be built?” Now it’s “can you describe what you actually want?”

The people getting the most out of these tools aren’t the best programmers. They’re the best delegators — the ones who can hold a clear intent, hand it off, inspect the result, and correct course without micromanaging. That’s a management skill, not an engineering one.

What actually changed

Three things had to line up. The models got good enough to hold a whole small app in their head. The feedback loop got tight enough that a wrong turn costs seconds, not hours. And the interface got conversational enough that describing beats specifying.

When those three land at once, the cost of trying an idea collapses. And when the cost of trying collapses, the right move is to try far more things — which is exactly what this desk is: a pile of small, cheap, finished experiments.

So the tools are ready. The remaining question is whether we’re willing to change how we think about our own work — from doing, to describing, deciding, and directing. That’s the harder upgrade, and it’s the only one left.