rethink your priors

ep. 2 · · ai implementation

I cancelled my apps. Then I killed mine.

"I cancelled my subscriptions. Then I killed my own app."

the prior it challenges:

You need an app for that

Transcript

Rethink your priors.

Over the last four months, I cancelled my subscriptions. Monarch Money. Kubera. MacroFactor.

And then I killed my own app.

Quick backstory. About three years ago I built an app called Clueless Clothing.

The whole reason it existed? I really don't know how to dress myself.

So I kept texting my wife and my sister, "Hey, does this outfit work?" The app replaced the texts.

That started me down the path of building apps for the App Store. Minor success.

And by minor, I mean my apps make about five or six hundred dollars a month. Which pays my A.I. bills. And I'm happy with that.

Then agents showed up. And I realized I could stop paying for Monarch Money, because I built my own household ledger.

It's a little crazy. A little intense. A lot more detailed. But that's how I am with data.

And it's convenient. I know exactly how much I spent on waffles this year. Not at one store. Across all the stores.

I can get down to how much I spent on blueberry waffles.

Because I take a picture of every single receipt. And the stores where I type in my phone number? They track what I buy. So I pull my history, and now I have it too.

Kubera did my long-term financial planning. Now my CFO agent does that too.

This morning I did the same thing to MacroFactor. I paid for it for years to track my meals.

I exported my data, about three years of it, and handed it to my agent. "Here's everything. Track my meals."

It came back with recommendations, and basically an algorithm we could use. I said, that's perfect.

Then I gave it to my food steward. That's the agent that handles all our meal planning and meal prep.

MacroFactor never talked to my meal planning Google Sheet. Now it talks to all of it.

Plus the household ledger, which my CFO agent runs. So the food steward gets its budget from the CFO.

And when I send it a recipe I saved, like, "Hey, this is a cool meal," it already knows to adjust it. I don't have to tell it.

And then I realized something.

(kid yelling) Daddy.

Why am I expecting people to pay for Clueless Clothing, when I just cancelled all of these other subscriptions? I have to kill this thing too.

So Clueless Clothing goes away as an app. It comes back as an agent version. Maybe a connector. It can keep track of my clothes, the weather, all of it.

My take? Tracking your clothes, keeping them in a database, that should be free.

Because what people were paying for, me included, was reasoning on top of the data.

Now the agent provides that. Grok Bot, Muse, whatever you use.

So what I need to sell you is something else. I don't know what that is yet. But I know there's value beyond what the app was doing.

And if you want to try this yourself, don't start by building a database. Start with a Google Sheet. Have your agent use the sheet.

See if you like the system. See if you like the outputs. It won't be efficient. It won't be ideal. That's fine. It works.

Start small. Get it to where you like it. Then scale.

Anyway. Start from the outcome. Or don't. I'm not your manager. Follow for more.

Priors from this episode