Because writing code was never the whole job. Understanding systems, identifying risks, designing coverage, validating AI-generated tests, debugging failures, and owning quality still require engineering judgment.
AI may change who writes the automation. It doesn't eliminate the need for someone who knows what should be automated, why, and whether the results can be trusted.
The real shift is from writing scripts to engineering quality.
~ @beinghumantester What we say for coding same definition goes for scripting. Scoping is something that needs human judgement and even if ai can do whole automation the problem lies in accountability that who will be accountable if something is missed.