These notes are very old, as I haven't been able to write them for a while. Some of them do not make sense to me now, so here's what I remember for this week.

• I've been helping as some sort of technical advisor for hiring, and I've learnt a lot about people and evaluation. I also learnt about myself a lot. I consider myself a jack of all trades, and a master of none. But I've learnt to understant that there are some trades that are not something I want to learn right now. Most of the times, salespersons, and political roles are what make or break organizations, both outwardly and inwardly. Finding a good fit is hard to find and it is not a technical problem.
• I found this https://sudaiv.net/blog/opencode-with-lm-studio-local-llm/ and I've been thinking to move my LLM workflows to the K3s cluster I built years ago. I realize I can't wrap my head around building applications on top of a system that is hardly deterministic.
• This told me my old laptop nodes are very limited to run models https://github.com/Pavelevich/llm-checker
• I also found thanks to https://replacebyclawd.com/ that I'm highly replaceable. This scares me but in turn it collides with the idea that working and teams is ultimately a people problem that won't be solved by AIs.
• A friend visited and I got to build some paper planes with him, I have the World Record Paper Airplane Book at home. We went out at night to fly those paper planes that fly in a loop, and they really turn! This book came to mind, which I found in the next weeks after this.
• I got a cold this week, so I stopped running, but rode my bike.
• I also finished my terraform slop project, learned a lot about GCP, terraform, IAM, networking. I used AI to generate hands-on exercises to understand trade-offs and subtleties about managing systems on GCP. It's really hard to sit and watch videos for the certification, I rather build things myself.