• I read this post and I feel Github and/or Gitlab have better tools to help maintainers filter out slop from codebases. More than that, I feel we as developers (or hackers, if you want) owe it to the rest to be transparent when we use AI. Add agents as collaborators, say it explicitly in commit messages, etc.
• I also read about a prompt injection making AI install AI in this post and I sort of admire the efforts from both the attackers and the defenders.
• I am using AI to get some hands-on experience using Iceberg, Nessie and Marquez. I'm building content that adapts to my learning style, and I'm about to finish it.
• There's a saying in Spain I've heard when talking about planning (I've only ever heard it in Spain, so I can't say for sure it's a thing in other Spanish-speaking countries): "estimar es timar". It roughly translates to "To make estimations is to scam". In Catalan, "estimar" can be used to say "to estimate" but it also means "to love", so the saying goes more like "To love is to scam".
• At the beginning of 2025 I wanted to pursue a PhD in Knowledge Graphs and their implementations in LLMs. The field moves really fast, and I don't, and I ran out of steam quickly. AI services have improved tons since that time and I can build POCs very quickly. I want to give it a go again once I'm done with learning GCP and modern data architectures.