Alice Yu

Every shortcut removes a bit of friction, and some of that friction was doing a job. I'd like to hear which parts of your workflow you handed over and which you took back.

When a tool makes something easy, you rarely notice what the difficulty used to give you. Writing a first draft by hand forced me to decide what I actually thought. Reading user feedback line by line made me notice phrasing that a summary flattens. Writing a spec from scratch was slow, and the slowness was where I found the holes.

Every user has different opinions about what your product should do

Some want new features, some want a better design, and others want things that might not fit your product at all

Trying to satisfy everyone can make things confusing

I think the challenge is understanding which feedback points to a real problem shared by many users

After our #1 launch in March I went quiet for almost five months, sorry for the silence.
When I came back, my old way of building with AI stopped working.

My Claude and Codex limits were gone in 2-3 days. Codex kept forgetting what the other sessions had done. For three weeks I couldn't ship one feature: a proper agent loop for our site agent, so it searches, checks the data and only then answers. I kept forwarding messages between the two AIs by hand, like a courier.

In the end I set up a harness: one shared space where every AI session, in any tool and on any computer, starts by knowing what we decided, what we tried and what's next. Within two days the stuck feature finally worked. Now my limits last the whole week.

I wanted to share this experience and discuss it here. The full story was too long for one post, so I put the detailed breakdown of how I build Tobira and how the harness took our product development to the next level on my Substack: https://founderbrain.substack.co...