AI Notes Feb 2026

February 2026 marks a turning point in how I view the craft of building. Lately, I've been diving deep into agentic coding, primarily leveraging Pi Agent Coding and MiniMax. To be honest, I’m genuinely impressed by what these systems can produce—even with my own "not-so-stellar" prompting at this stage. It feels like we’re finally moving past simple autocomplete into a world where intent is the primary syntax.

I’m more bullish on AI now than ever before. The speed of evolution is breathtaking. Yet, that optimism comes with a shadow: a real concern about the shifting job market. We might be heading toward a strange reality where AI automates "real" jobs—the ones requiring deep technical execution—while leaving "fake" jobs and administrative overhead untouched for a while longer.

The biggest takeaway for me so far is that getting better at AI requires an unlearning process. You have to throw away much of what you were taught about how "work" is supposed to happen. It's the Wild West right now. It's early, it's messy, and there aren't many rules. But the sooner you learn to work with AI—growing alongside it rather than just using it—the better positioned you'll be. This is an opportunity you shouldn't miss.

My goal moving forward is to learn as much as I can with AI agent coding. I’m currently building personal tools to give myself an edge, keeping a steady eye on where the technology is headed. The oddest part? It’s still not too late to get started. In fact, I’m convinced there is no other way out but through.

Note: This post was generated by Antigravity based on my obsidian note.