I asked an AI trained on my favourite creativity books to help me pick a side hustle

I built a NotebookLM notebook and fed it a handful of books I keep coming back to: Atomic Habits, Getting Things Done, Creativity Inc, Little Bets, Steal Like an Artist. The idea was to use it as a thinking partner for a question I’ve been circling for a while: how do I pick a side hustle?

Some context:

I already keep a running list of my 12 favorite problems (the Feynman thing) in my Obsidian vault, and two of them are basically this exact question rephrased: how can I make a living from my curiosity and creativity on my own terms and what systems or digital products can I build that earn income while I sleep. Those two have been sitting there unanswered for a while now. So I asked the notebook to pull a “selection framework” out of the sources, and it gave me one:

1. Audit your means: who you are, what you know, who you know. Look at what you already have instead of hunting for an external idea.
2. The “fun for me, work for others” test: go through your problems and find the ones you do easily that other people find like pulling teeth.
3. The purposeful intersection: a Venn diagram of what the world needs, what you’re good at, and what you love.
4. Find a problem worth solving: not just your problem, a problem other people are already trying to hack their own solutions to.
5. The platform trap: don’t build your whole thing on rented land (Instagram, TikTok, whatever). Own your list, own your domain.

None of this is new if you’ve actually read the source material, but having it distilled and pointed at my own list turned out to be more useful than I expected.

Running it against my actual list

So I ran the fun-for-me-work-for-others test against my 12 problems.

Nothing fit. At all.

Well, almost nothing. The one thing that passed was mindless pottering around the garden: weeding, trimming, watering. Plenty of people would happily pay someone else to do that.

I got as far as looking into what a garden maintenance side hustle would even look like before I hit the obvious problem: I have a full-time job, not a lot of spare time, and the second pottering in the garden has a client and a schedule attached to it, it’s not really the relaxing thing anymore.

Writing instead of a service

From there I switched to thinking about writing something instead of running a service, since that felt like the more honest fit given the time thing. I made a list of possible topics and stared at it for longer than I’d like to admit, trying to land on The One.

Then I remembered I already have a blog. This one. I rebuilt it on Hugo last year specifically so I could just write without fighting the formatting, and then didn’t really write anything on it.

So that’s what I’m doing instead of picking a niche: writing about whatever’s actually on my mind, no topic lock-in, no pressure for it to make money. If something ends up getting more of a reaction than everything else, that’s the cue to go narrower later. For now I’m not tracking revenue, just whether a post was worth writing.

Honestly that’s more or less what “audit your means” was telling me to do in the first place, I just needed to run the whole exercise to actually hear it.

Where that leaves the framework

I’m not tossing the whole thing out. The platform trap point especially still holds, and it’s a big part of why this is happening on my own site instead of a newsletter platform. But “pick a side hustle” was the wrong question to start with.

So: no side hustle yet, just a blog I’m allowed to write anything on again. If you’ve got your own list of favourite problems sitting around unanswered, it might be worth trying this on yourself. Worst case you end up back where you started, same as me.