# I pulled together a working Alpha 1 of Larc Mini about a month ago. Now I'm unsure of what next. ## Grand Vision & One Loyal User Driven by the grand vision of a "learning architect", I dreamed of a personal AI agent integrated into the note-taking space that can guide one through the learning process and provide support. I eliminated all the agentic features and built the bare minimum -- an AI Chat Sidebar that has the context of the currently open note, with quick activation for custom prompts (skills). It's just Obsidian + Copilot, pulled together and packaged. I think it's working fine, and there must be use cases for it. But it's just I think -- I have to let people actually test it and see if they will use it. The idea is that even if I can only get one person in our communiy -- not very technologically progressive in general -- to use it, there is a point to keep developing. Just one user I need to validate my idea. ## The Urge to Share I don't think I am someone who would sprint into a lecture hall at the end of a lesson to cold-pitch my product, like the CEO of Cluely. But I might really need that energy to at least market for the first few users. Just as in the cold start problem, one should take whatever measure, even unscalable ones, to get the first community. So I might really need more of the urge to share my work and push it to the world. ## What's the point? Even if I can find a long-time user that develops dependency on Larc Mini, how will I be able to distribute it to the wider public is still going to be a big challenge for me, especially as a first time developer. Besides techinical difficulty such as removing my hardcoded Deepseek API, providing BYOK and model selection options, correctly notarizing it, figuring out how to push updates, etc., I imagine lots of challenges from a reddit community for example -- should I opensource it? how do I pitch it? what's the point of pushing it to the raw market, if there is not a business model, and it's hard to get valuable feedback (I have no built in tracking)? ## The Realization And then the realization hit upon me -- it might be the case that the entire point of sharing is just to find that one guy, who has more experience than me, and also buys the idea / vision, who will actively contact me and provide help / guidance. Maybe that person is whom I truly need, even more than the first loyal user. ## Human expertise & experience still matter AI did raise the floor and made it exponentially easier to build a software -- but again as the famous diagram illustrates, it can only take you thus far. Above a certain level, human expertise and experience still matter -- because that's what determin if a product can survive when pushed the the test of the true market, beyond MVPs and ideation. ![[Screenshot 20251015 at [email protected]]] [source: I is a Floor Raiser, not a Ceiling Raiser](https://elroy.bot/blog/2025/07/29/ai-is-a-floor-raiser-not-a-ceiling-raiser.html#coding-a-boon-to-management-less-so-for-large-code-bases)