Stop Building a Productivity System for Someone You’re Not

You know that pair of jeans in the back of your closet? The ones you bought because they were going to fit once you got in better shape?

You never wear them.

That’s how most people build productivity systems. They design a routine for who they want to be, like a highly disciplined morning routine, then wonder why the whole thing sits unused. Same as those jeans. Does this sound familiar?

I compare this to buying clothes that fit a future, skinnier version of yourself. That’s the Ideal Self Trap. You build from the person you plan to become instead of the person who has to live inside the system today. If you’re a parent with young kids, you already know how fast an ideal routine falls apart. Systems have to start from your current worst part and get iterated upward, or they don’t last.

AI made this trap easier to fall into. You can dress the Ideal Self in even fancier clothes now.

Your five-phase workflow just collapsed

If you’ve followed GTD, you know the traditional workflow: Capture, Clarify, Organize, Reflect, Engage. Five phases. Processing used to be the human bottleneck. You captured a pile of stuff, then you sat down and sorted it by hand.

That workflow has collapsed into a single step.

You capture the intent. AI handles the processing and the organization in seconds. The human role has shifted to defining the outcome and reviewing the final quality.

I used to treat my inbox like a list of tasks. Now I treat it like a stream of intent that AI organizes. The shift happened when I stopped typing as the default and started talking. I’ve dictated over 1 million words using Whisperflow. A 400-day streak. Voice capture is superior to typing for a lot of what I do because it’s immediate and low-friction, and AI can process what I dump out. I don’t have to Clarify and Organize by hand the way I used to.

Have you tried capturing by voice for even one day?

Try it. One day of dictation. Watch what happens to your capture habits, then pick one manual processing step in your current workflow and hand it to AI this week. Capture stays human. Processing doesn’t have to.

The Outer 10% is now the job

With AI handling 80% of the work (processing, organizing, executing), your critical role is the outer 10%. Two moves. Define a clear outcome as the input. Review the quality of the output.

The hardest part is no longer doing the work. It’s reviewing it.

AI can produce volume. That volume still has to meet your standard, and AI will not do that check for you. Review is the new high-value skill.

Spend 10 minutes this week on an AI-generated output. Read it critically. Does it match the outcome you defined? If it doesn’t, that’s the work now.

Brooks over-prompted an agent. It got slower.

Brooks was trying to be helpful. He gave an AI agent detailed, step-by-step instructions to download videos from a platform. The agent followed the slow, manual path he laid out.

Later he used a simpler, outcome-based prompt. Same goal. The agent discovered a faster API method it already had access to.

Over-prescribing the steps blocked a better solution.

There’s a difference between training an AI and delegating to it. A tutorial locks it into your current method. An outcome lets it find a route you wouldn’t have taken.

Give an AI a high-level goal this week instead of a step-by-step tutorial. See if it finds a faster route.

Stop shopping for the perfect app

Brooks has another confession a lot of knowledge workers will recognize. He spent a long stretch jumping between writing tools (Pages, Scrivener, Ulysses) looking for the perfect environment. That shopping hindered actual output.

The core fundamentals of the work stay the same. AI should enhance the workflow you already have. You still have to sit down and write.

This is the Ideal Self Trap in a different outfit. The next app is the pair of jeans.

Identify one tool you’ve been shopping for. Commit to your current setup for one week. Optimize the workflow, not the software.

Brooks also recommends using meme generators in Slack or Teams, specifically the midwit bell curve meme, when a discussion gets over-complicated. It’s a cultural shorthand. You drop it, people come back to the simple version. If you lead a team, try visual shorthand the next time a conversation starts spiraling.

Run your AI tools through TEA

The TEA Framework (Time, Energy, Attention) is still the most relevant lens I have for this. It’s abstract enough that it holds regardless of which tools you pick.

AI primarily saves Time by automating drudge work. It can aid Attention if you’re using it for daily briefs and dashboards. Energy is a toss-up. AI can reduce the drudge. It can also lead to overwork if you never set a boundary.

Look at your current AI tools through TEA. Which one saves you the most time? Which one drains your energy? Keep the first. Put a boundary on the second.

Audit your current system for Ideal Self elements. The morning routine that only works on a perfect day. The processing you still do by hand. The tool you keep meaning to switch to. Circle the ones causing friction. Rebuild those pieces from your worst day, not your best one. Then iterate upward.

That’s how you get a system you’ll actually wear.