• This week is shorter because I'm going on vacations today afternoon.
• This week started with some challenges. I need to read more about leadership strategies, how to get more credibility and buy-in. This week also I talked to some people that are going to help me with these things, I hope.
• It's hard to think of one's career from a non-capitalist point of view.
• I finally started playing with MCP servers after a lot of resistance from my side. Since I use copilot I need to adapt most Claude examples to my scenario. I see its power and its value, I still need to get an idea about how to audit these things.
• Particularly about the recent developments of moltbot is that competence is more important than ever.
• I read code and I sort of understand now, but still I'm running most agents in sandboxes. I can't let the agent perform actions without prompting me first. My guess is that most MCP-enjoyers have some sort of guardrails in place I'm yet to discover.
• I've been reading lots about developments in AI and the Davos forecasts are worth reading. The future looks fascinating and dystopic at the same time, is this what always happens, like, with our previous generations? What will happen with the next generations?
• I've seen some repositories of MCP servers featuring a badge with an unnamed animal on them. Like a frog, an insect, etc. I guess it started with the O'Reilly book covers and it's a common trend now, but it is moving in a way that computer things try to engage with people asserting some kind of bond with nature. There's nothing less engaging than reading the MCP specs, unless you are a nerd.
• I've been watching Task and it is really good. My guess it's about people having to make difficult decisions and there's no way to tell as a viewer what the best outcomes are. It makes me really anxious.
• I also read this note about leaving management tasks to Claude from Bernardo, we worked together in a group project for the Full Stack Deep Learning course and I'm glad to see he is thriving.