AGI is the elephant in the room. Whether it's one year or ten years away, the trajectory is the same: the "manual labor" of coding is disappearing. Our job is evolving into being "System Architects" or "Product Engineers" rather than just "Coders."

It’s a chain reaction! Managers are under pressure to deliver more with smaller teams because "AI makes it easy," which in turn makes them hire fewer juniors. It’s a tough cycle, which is why being a high-signal junior is so important right now.

FAANG (or MANNGA) has doubled down on fundamentals. They know you can use AI for boilerplate, so their interviews are leaning even harder into deep systems knowledge, low-level optimization, and how you handle ambiguity—things a prompt can't solve yet.

Open source and Linux are actually the best "training grounds" left. If you can contribute to a major open-source repo, you’re proving you can read, understand, and improve complex code written by others. That is 100x more valuable to a recruiter than another solo project.

Well said. Being a "code archaeologist"—digging through old, messy files and figuring out why they’re breaking—is a superpower in 2026. Anyone can start a project; very few can fix a dying one.

Spot on. The market isn't necessarily "dead," it's just filtered. The "cold" market is mostly for people who stopped learning at the syntax level. If you can speak the language of system design, there’s still a lot of opportunity.

That "validating logic" shift is huge. We’re moving from being the "writer" to being the "editor-in-chief." You have to be smart enough to catch the AI's subtle mistakes, or you’ll just end up shipping faster bugs.

It’s the only way to stay relevant. Boilerplate is now a solved problem. Solving the puzzles that arise after the boilerplate is where the paycheck is.

The shift is permanent. "Doing" is a click away for everyone now. "Understanding" is still something you have to earn through hours of staring at a debugger. That's the real moat for developers today.