Technology Judgement
When Building Becomes Cheap, Judgement Becomes Expensive
AI increasingly helps produce the code, architectures, tests, analysis, options and documentation that used to take a team weeks. The scarce capability shifts as a result. The interesting question stops being "can we build this?" and starts being "should we?"
That's a judgement question, not a technical one. It depends on context, trade-offs, evidence, and — just as often — knowing when not to act at all. None of that gets easier just because production gets faster. If anything, it gets more expensive to skip.
This is where the thread I've been pulling on across the last two notes ends up, for now: execution accelerating, decisions needing to become more reversible, and underneath both of those, judgement quietly becoming the thing worth paying for. It's a question this book keeps returning to, in different clothes each time.