A new era for software testing
Summary
The article argues that automatic programming and large language models can accelerate software development in certain contexts, but may not reach the efficiency and elegance of hand-written code. It highlights a specific way AI can enhance software QA by acting as a QA engineer to perform automated, guided testing across new releases, using markdown-driven test plans and remote endpoints. The author suggests that automatic QA could raise the bar for release quality and compensate for faster, AI-assisted development, though it requires careful management to avoid regressions and missed edge cases.