The intelligence-to-judgement ratio determines which professions AI automates first

Bek draws a sharp line between intelligence and judgement: “Writing code is mostly intelligence. Knowing what to build next is judgement.” Intelligence work — translating specs into code, medical billing, contract drafting — follows complex rules, but they are rules. Judgement requires experience and taste, instinct built on years of practice. Software engineering crossed the automation threshold first: “more tasks are started by agents than by humans” and it “accounts for over half of all AI tool usage across professions. Every other category is still in single digits.”

The implication is that the intelligence-to-judgement ratio is the key predictor of when any profession faces disruption. This connects to Production agents route routine cases through decision trees, reserving humans for complexity — production systems already separate intelligence routing (deterministic) from judgement calls (human escalation). Today’s judgement will become tomorrow’s intelligence as AI systems accumulate proprietary data about what good judgement looks like, shifting the frontier continuously. Crucially, automation doesn’t eliminate the human side — AI automation amplifies demand for expert human judgment rather than replacing it shows that while pre-labeling cut simple task costs 100,000x, demand for expert judgment at $200/hour actually grew. Alfred Lin’s observation that AI compresses the distance between idea and execution but not between good and bad judgment extends this from professions to individual practice: within software engineering, AI compresses the execution layer (coding, testing, deployment) but leaves the judgment layer (what to build, what to stop, when to pivot) as the binding constraint. When When production constraints dissolve, the bottleneck shifts from execution to judgment, the intelligence-to-judgment ratio becomes not just a predictor of automation but the primary axis of individual competitive advantage.