Ask HN: What are your best ideas you will probably never build?
I think a lot of people here have ideas they keep thinking about, but realistically will probably never build.
Maybe they are too small for a startup, too weird for work, too niche, too much effort, or just too far down the backlog.
I wonder if it would be fun to have a thread where people pitch those ideas.
If you have something you would like to see exist, but probably will not build yourself, post it here.
Could be software, hardware, data, art, local tools, research, weird experiments, anything.
I'm curious about your ideas! Use https://github.com/ndjordjevic/pin-llm-wiki as scaffolding (or a similar project to automate Karpathy's LLM Wiki pattern) to support keeping a wiki continually updated containing an HN user's comments updated as a Wiki knowledgebase, with the option to annotate or correct facts and similar context/knowledge accordingly. Automated knowledge base of sorts, grounded with the target user's comments (or multiple users, depending on whose comments you value). I want to build a modern version of a code book. Something that uses zero encryption at all, yet contains everything I could ever want to say. It seems like once encryption took over the art of codebooks kinda went away but I still find the history of them really interesting and think a modern take would be fun to build. its a programming language, where code or its execution just behaves different depending on the case of use . a context aware compiler probably analyzes the environment, then it compiles in different ways. could also be hardware attached How about yours? if you start recording your whole day including what you said/heard/seen you can build your persona LLM and leave it to your kids Meetup alternative.