Journal · The AI Architect · 2026-07-21
A task is something you do once; a tool is something you run again. Do the thinking once, then just press go.
A task is something you do once; a tool is something you run again. Do the thinking once, then just press go.
The real tension isn’t between laziness and effort. It’s that thinking and doing feel like the same motion, so we never separate them. You sit down to write the weekly report, and you start composing sentences immediately, as if the only way to do the task is to be inside it. Nobody teaches you to step back and ask: is this a one-time thing, or is this the third time this month I’ve done exactly this?
That question is the whole game. Most people don’t lack discipline. They lack the pause where they’d notice a pattern repeating.
Notice the rerun before you do it again
A task feels urgent and singular even when it isn’t. The report due Friday feels like its own event, disconnected from the report due last Friday. But if you’ve done something more than twice, in roughly the same shape, you’re not looking at a task anymore. You’re looking at a tool that hasn’t been built yet.
The habit worth building isn’t “build more tools.” It’s noticing repetition while you’re in it, not after. A useful signal: if you catch yourself thinking “I did this before, how did I do it last time,” that’s the moment. Not next month. Right then.
Try this: for one week, keep a small list — three columns, nothing fancy — of anything you do that you’ve clearly done before. Just note it. Don’t fix anything yet. At the end of the week, look at the list. You’ll usually find two or three things that show up more than once, and seeing them written down does something that noticing them in passing doesn’t. It makes the repetition undeniable.
Do the thinking when you’re calm, not when you’re rushed
Here’s the part people skip: building the tool has to happen when you’re not under the deadline that made you need it. If you try to design the system on the same day the report is due, you’ll just do the task again, a little more self-consciously, and call it progress.
The thinking — what are the steps, what changes each time versus what stays the same, what decisions actually require your judgment versus which ones are just habit — needs its own separate time. Ten minutes on a Tuesday when nothing is on fire is worth more than an hour on a Friday when everything is.
This is why so many good intentions to “systematize things” fail. People try to do the design work and the deadline work in the same sitting. The deadline always wins, because it’s louder.
Try this: pick one recurring task you identified. Block twenty minutes this week, not to do the task, but to write down its steps as if explaining them to someone new. Where do you pause and think? Where do you just repeat the same move? The second kind is what becomes the tool. The first kind stays yours.
Build the tool smaller than you think you need to
Once people do decide to build something reusable, they often overbuild it. A checklist becomes a twelve-step process. A simple template becomes a system with branches and exceptions for every edge case they’ve ever hit. The tool ends up harder to run than the task was to just do by hand.
The tools that actually get used again are almost always plainer than they feel like they should be. A template with three fields. A five-line checklist. A short list of questions you ask yourself before starting, so you’re not re-deciding the same things every time.
The test isn’t “does this cover every case.” It’s “can I press go without thinking.” If using the tool requires as much judgment as doing the task from scratch, you haven’t built a tool. You’ve built a more elaborate task.
Try this: when you draft your first version of a checklist or template, cut it by a third before you use it the first time. You can always add back something you actually missed. It’s much harder to strip out steps you’ve grown attached to.
None of this is complicated, which is exactly why it’s easy to skip. Noticing repetition, protecting a little calm time to think it through, keeping the result small enough to actually use — that’s most of it. The book goes further into how this scales when the repeating task involves other people, or when the “tool” is something closer to a habit of mind than a document. But the shape of it starts here, with paying attention to what you’re doing a second time.
Go deeper. The full method is in The AI Architect. New here? Start with the free companion pack, or explore the series.