Last time I gave you a ladder with four rungs and asked you to write down yours. Most of you wrote a two. That’s fine. This episode is the first step off it, and it has nothing to do with picking a tool.
If you missed Episode 1, it’s here: Tools Are Rented. Workflows Are Owned.
The question everyone starts with
Where could I use AI?
It sounds like a good question. It isn’t. When you ask it, you go looking for a place to put a tool. You open the tool, you poke around, you find something it can do, and then you try to make that thing matter. The tool is choosing your problem for you.
That’s not workflow design. It’s tool-hunting in disguise.
Start with the tool and the tool picks your problem.
The question that works
What do I do over and over that costs more than it should?
There’s no AI in that sentence. You’re not looking for a tool. You’re looking for a job. The tool comes last, and by then the job has already told you what it needs.
Start with the job and the tool becomes a detail.
Three tests
A task is worth building a workflow for only if all three of these are true. Not two. All three.
- Recurring. You’ll do it again within a month. If it’s once a year, just do it. A workflow for a once-a-year task costs more than the task.
- Structured. It has steps. If it’s pure improvisation, that’s a craft, not a workflow, and you should keep it.
- Draining. It costs you something you’d rather spend somewhere else. Time, attention, patience, Sunday night.
Run some real tasks through it and the tests start doing their job.
The annual grant application. Everyone hates it. It happens once a year. Fails the first test. Just write it.
The hard conversation with a parent. Recurring, unfortunately. Draining, absolutely. But it has no steps. Every one is different. That’s a craft, and it’s yours.
Sorting Friday’s exit tickets into who needs reteaching and on what. Every week. Same steps every time. It eats an hour you’d rather spend with the kids who need the reteaching. That’s the one.
Three tests, and the task that passes all three is rarely the one you’d have guessed.
The test people skip
Draining is the test people skip. They find something recurring and structured, they build it, and they never open it again.
Draining is the only test that predicts whether you’ll still care in a month. If the task doesn’t cost you anything, you won’t maintain the thing that fixes it.
If it doesn’t hurt, you won’t keep it running.
Boring wins
One more thing before you pick. The tasks that sound exciting are usually the wrong ones. An AI lesson generator. A chatbot for students. Something that grades essays. An AI tutor for every kid.
The tasks that sound boring are the ones that work. Friday exit tickets. Sub plans. The weekly family update. Meeting notes into a tracker. Boring is what you’ll actually run next Friday.
Boring workflows are the ones that survive.
Your homework
One sentence.
Every ___, I ___, and it costs me ___.
Mine: Every Friday, I sort exit tickets into three reteach groups, and it costs me an hour I’d rather spend with group three.
That sentence is the whole assignment. If you can’t fill the last blank, the task isn’t draining and you should pick another one.
Next episode: we take that sentence and map the real steps, including the one where you procrastinate.
The series:
THE SERIES Episode 1, Tools Are Rented. Workflows Are Owned.: https://youtu.be/AT3pVm4edRk
Full playlist: https://www.youtube.com/watch?v=AT3pVm4edRk&list=PLPPRhbfNqqyI
Episode 3 (Map It and Mark It): coming soon
What went in your last blank, and did it surprise you?
Leave a Reply