> And then... then what? It turned out that I never reached out to any of these notes. Most of the useful knowledge is already right there, in my brain. I don't

> And then... then what? It turned out that I never reached out to any of these notes. Most of the useful knowledge is already right there, in my brain. I don't need a second brain for that. This sort of approach shines when you need to mobilise knowledge that fell out of your immediately accessible memory, e.g. 6 months after having done something for the last time. It’s not very well suited to fields that move too fast and where you never need to look back, because by then the knowledge is outdated and you need to find new solutions anyway. For me, it’s great because I have a handful of long-running projects concurrently and I switch between them every 3-4 month. Being able to easily mobilise project-specific knowledge after having spent enough time doing something else that I forgot the details is very useful. To be honest, I would love a LLM that I could use to extract information from my old pile of notes and my pile of scientific papers. I don’t think it would make notes useless, though. On the contrary it would make them much more valuable.