Stop Carrying It All in Your Head

Recently, two coworkers asked me nearly the same question within a day of each other.

“Did you see the email I sent you two days ago?”

“Did you get a chance to answer my question?”

No, and no. When I finally sat down with each one, the answer took about ten minutes. The work wasn’t the problem. Finding the work was.

Messages come at me from every direction now: email, Teams, DevOps, text. Somewhere in that pile were two people waiting on me so they could do their jobs, and I didn’t even know they were there.

The way I used to work

For most of my career I was proud of keeping my to-do list in my head. I’d pull things out of thin air right when I needed to remember them, and I told people that’s how I worked best.

Fifteen years ago, that was mostly true. But I think I had the reason wrong. My world was smaller. I lived in my inbox all day. Email was essentially the only way to reach me, so everything came through one door, and nothing got missed.

That door is gone. There are five or six doors now, and I’ve been trying to stand at all of them at once.

The habit underneath it

Here’s the habit I’m actually having to break: I assume that if something needs doing, I have to be the one who gathers it, sorts it, and figures it out. Every single time.

For most of my life, that was just the price of getting anything done. You set time aside. You educated yourself. You researched the problem, diagnosed it, and came up with a plan of attack. As recently as a few years ago, there wasn’t another way.

Now, with AI, almost any question I can ask leads to a better next question, and the next one, until I land on “take these steps.” The research isn’t the bottleneck anymore.

So I’m learning to ask a different question: how do I set this up once, so I’m approving the work instead of rebuilding it every time?

The thinking doesn’t go away. It moves. I still have to sit down and think hard, but I do it once per system instead of once per problem.

What that looks like in my house

The one that’s already working for us is the weekly menu.

We can’t have soy or rice in our house. That rules out a lot of recipes, including a Vietnamese-style dish we call blazing noodles that we really love. Two of its primary ingredients were simply off the table. In the old days, I’d have guessed at substitutes and hoped for the best. This time I asked the tool, and it gave me alternatives that worked, quickly, without an evening of searching.

Planning the week used to take me 30 to 45 minutes: figuring out what we wanted, digging through recipes, converting the ones that didn’t fit. Now it takes about 15, because I’ve already done the hard part once. We have approved ingredients and approved recipes. The tool works inside those lines, and I sign off on the result.

I didn’t hand over the decision. I handed over the gathering.

What I’m building next

The messages problem is the one I’m working on now. The plan is a single set of instructions I run two or three times a day that pulls every channel into one view: what came in, what needs an answer from me, and what I can mark resolved. At work, that means using only the tools my company has approved for work data. Over time it becomes a log, so “did you see my email?” already has an answer before anyone has to ask.

And once I started thinking that way at work, I saw the same problem at home. Have the bills been paid? Do we have room for the next thing we want to do? Is there a subscription I should have canceled months ago? Right now those answers live in a dozen places. Seeing them on one screen, where I can tell how they tie together, would change how I make decisions.

Where this goes wrong

The temptation, once you’ve built an automation, is to trust it because you built it.

When you’re automating your communication with people, you have to be careful you aren’t over-filtering, and that something is actually reading your messages, whether that’s you or a tool that can tell when something matters. Even with the best intentions and the best setup, things get missed.

In my day job we never cut over to a new system without running it alongside the old one first. The same rule applies here. For a while, I’ll do it the old way and the new way and compare them. Only when the new way proves it won’t miss anything do I let the old way go.

An approver who doesn’t actually look is just a rubber stamp. The point is to spend my attention on the decisions instead of the digging.

If you’re uneasy

If you’re still a little uneasy about using AI in your everyday life, or quietly wondering whether it’s cheating, I get it. I had to admit to myself that it’s a tool, and used well, it keeps me focused on what I set out to do instead of buried in everything around it.

If I were starting over, I’d pick one thing I rebuild from scratch every week, write down how I want it done once, set it up, and check its work for a few weeks before I trusted it.

I like getting things done. I like my family more.

Leave a Reply