# I asked AI what's wrong with me

> 42 projects, one honest review, and the page I now read every morning.

Source: https://umairalikhatri.com/writing/i-asked-ai-whats-wrong-with-me/
Author: Umair Ali Khatri
Updated: 2026-11-18

---

[&larr; All writing](https://umairalikhatri.com/writing/)
Notes

# I asked AI what's wrong with me

18 November 2026 · 3 min read

I asked Claude to do a personality check on me. Not the nice version. The good and the bad.

We have worked together for a long time. It has seen my half-finished plans, my late-night ideas, my new projects. So I asked it to be honest.

It was.

## The good

I start things. A job, a design community, a small business, a perfume idea, client sites, animations. Most people talk about one side project. I run several.

I build people. I grew a design team from 2 to 16.

I want proof. I push back on claims, even when they come from AI.

I show instead of argue. When I doubted an estimate, I built the thing myself to see.

## The bad

Too many fronts. Strong start, then the next shiny thing.

I build systems around the work. Manuals, setups, templates. They feel productive. Past a point, they replace the work.

I do work my team should own. Proving a point once is fine. As a habit, it teaches people that I will step in.

My briefs are short. They make sense to me. Not always to the person reading them.

That last one stung, because it was true.

## The number

Then I looked at my Claude account. 42 projects.

42 open loops. Each one a small promise I made to myself and didn't keep.

My first move was to organize them. Groups, prefixes, new names. It looked clean.

Claude pointed out that this was the second bad habit. I had made the list tidier. I had not made it shorter.

Fair.

## The cut

So I set a rule: eight active projects. That is the limit.

Three for my job. Two for clients. One personal. Two ventures of my own.

Picking the two ventures was the hard part. Five felt alive. I picked a perfume line and my own writing and content. The other three went to the archive.

Archived is not dead. It is waiting its turn. But it is no longer on my screen every day, asking for attention.

42 became 8.

## The page

Then I wrote one page that I read every morning. Plain words. No dashboard.

Every morning, ten minutes, before Teams or email:

- Read the page.

- Pick the one thing that would make today a good day.

- Check that it belongs to one of the eight. If it doesn't, it is not today's thing.

- Give it 90 minutes before noon.

During the day, three questions:

- Before I start something new: is this one of my eight? If not, it goes on the Later list.

- Before I do it myself: could someone on my team do this 70% as well as I would? If yes, I hand it over.

- Before I send a brief: does it say what, why, what done looks like, and by when?

Every evening, five minutes: Did the one thing get done? Yes or no. What is tomorrow's one thing? Close the tabs.

Every Friday, twenty minutes: Did each of the eight move this week? Anything stuck for two weeks, keep it or archive it. Then read the Later list.

## The Later list

This is the part I need most.

Every new idea goes on a list and waits seven days. Most look less shiny by then.

If one still matters after a week, it can come in. Only if something goes out.

## What I learned

New things feel like progress. Usually they are a way to avoid the hard part of the old thing.

Organizing feels like progress too. It isn't.

Deciding is.

If you have 42 open loops too, say hello.
[All writing](https://umairalikhatri.com/writing/)
