What I’m learning out loud.
Evergreen essays written in the language people search for: the real questions that come up while building products, systems, brands, and AI-supported work.
One profile can become many views
The profile needs one source of truth before every site gets its own version.
The profile needs one source of truth before every site gets its own version.
A product can be any format
Can a product be a book, store, course, or content lane?
App, book, store shelf, content system, SaaS, digital product. The format is not the point.
Books and content count
What counts as a product if it is not an app?
A shippable shelf is still a product.
The workflow has to be clear
What makes a workflow clear enough to ship?
The system ships when the path through it makes sense.
One product gets cleaner faster
Why does one product get cleaner faster than five?
Five half-finished ideas do not create five times the progress.
Needing to ship two is reality
How do I choose what to ship when everything feels important?
Wanting to ship five is usually noise.
Close is not shipped
How do I know when close is still not shipped?
Almost done is still not done.
Streamo.gg
What does building Streamo teach about AI, infrastructure, and taste?
What building a platform with AI tools actually taught me about infrastructure
If you can't figure it out, oh well. Moving on.
How do I stop over-organizing AI chats and move on?
Sunk cost is the ultimate speed killer. The minute a feature requires a committee or endless troubleshooting to survive, it gets cut.
Reduce the noise
How do I become the person who reduces noise at work?
Be the person who removes drama, not the one who brings it.
The container
How do I hold multiple brands without forcing them into one box?
Stop forcing an entire ecosystem into a single box.
Day 1 is just showing up
What should Day 1 actually accomplish?
The annoying dignity of starting before it looks impressive.