Week One

So last week I quit my job to become a solo developer/entrepreneur. Huge.

Things that happened in this very first week:

Tiny Projects

I’m kind of modelling my new life as a solo developer on this guy. He launched a bunch of small projects, and seems to have had a lot of fun along the way. This feels like a good approach.

Tragically, he seems to have stopped doing his weekly blog now that one of his projects - PromptBase - seems to have taken off (it even has its own subreddit).

So here is my goal:

launch one new project every month and see which (if any) make money.


I made good progress on my first project: todobot.ai. Please try it out (go easy on me though, it’s super janky - what’s the stage before alpha? omega?).

It’s a todo list that is backed by a large-language model (aka GPT-4). The initial product will:

  • rewrite vague tasks so that they are more actionable
  • break tasks down into smaller subtasks

I think a lot of todo lists fail because people feel no social pressure to do the things on the list.

For example, here’s my todo list (on todoist.com):

A sorry bunch of neglected tasks

What a sorry bunch of tasks. Hard to action, no deadlines, no subtasks. Is it any wonder that I haven’t made any progress on any of them for months?

Would I make more progress on the things that are important to me if I had a robot companion to help me? Let’s find out.

New Tech

TodoBot is built with a number of few technologies I’ve not worked with before, namely:

They both seem like good choices. They fall into the bucket of: “maybe not super slick or customisable but good enough and quite easy to use for small teams and individuals”.

However, I haven’t worked much with them before, so that slowed me down a little bit.

See you next week!

Please don't sign up to my newsletter