Manual QA is Dead

The Distinction That Matters

The episode opens with a deliberately provocative claim: manual QA is dead, and that's a good thing. But the host immediately refines this. What is actually dying is not testing—it is manual quality control, specifically inspection-based work that can and should be automated. The real problem is that many testers have been doing quality control while calling it testing, and companies are now automating that work out from under them.

The host draws a sharp line between two activities that are routinely conflated. Quality control is inspection: you know what you are looking for, you know where to find it, and you know what the acceptable tolerance is. Can the user log in? Does the system reject an underage applicant? These are checks with predetermined answers. They are gatekeeping mechanisms—if the check fails, you reject the build. This kind of work is increasingly automated through continuous integration, unit testing, and behavior-driven development. The host argues that this is not testing. Testing, properly understood, is about providing information that helps people evaluate quality. It is not a control mechanism. It produces information that might influence a decision, but the decision itself belongs to someone else. Testing explores unknowns: edge cases, emergent behavior, conflicts between requirements, risks that n