Between career break, funemployment and contemplating early(-ish) retirement. May consider getting back in action for interesting, meaningful opportunities that don't revolve around coding agents. Potential exception: initiatives that treat agentic coding as a liability to be contained, trying to address issues including but not limited to:

• Code Quality & Technical Debt: working on tools and practices that keep AI-generated code honest, based on the radical premise that working software should also be maintainable by humans.

• Cognitive & Comprehension Debt: bridging the gap between "the code seems to work" and "nobody beyond perhaps the one who prompted it knows why and how", without reverse-engineering intent from reams of auto-generated slop that was outdated by the next commit.

• Review Infrastructure: designing workflows that preserve genuine human oversight, as opposed to an engineer's reflexive "LGTM" after scrolling over a 4000-line PR for 10 seconds.

• Test Integrity: ensuring that "tests pass" remains a strong signal of confidence rather than security theater played by agents adding dummy mocks or rewriting failing tests to match broken behavior.

• Engineering Culture & Developer Experience: tracking the cost of top-down AI adoption mandates for the people doing the work, because that's nobody's official problem yet.