A new era for software testing
The article discusses using LLM-based AI agents for software QA and testing, complementing traditional test suites. By having AI agents manually inspect code changes, run integration tests, and evaluate user-facing quality issues described in a markdown file, the author argues that automatic QA can raise software release quality and offset quality losses from rapid AI-generated code.
Background
The author, **Salvatore Sanfilippo (antirez)**, created and long-maintained **Redis** — one of the most widely used databases in the world. He is respected for his emphasis on clean, simple code. His take matters because he's a veteran who actually builds production software.
"Automatic programming" means generating code with large language models (LLMs) like ChatGPT. The article's core claim: LLMs are best used not for writing code, but for **automated QA** — acting as a tireless manual tester that examines new commits, hunts for regressions, runs integration scenarios (e.g., distributed inference across two laptops), and even evaluates whether features feel user-friendly. This is a new approach on top of, not replacing, traditional unit/integration tests. Sanfilippo argues this could meaningfully raise release quality, especially as AI-generated code inevitably becomes more common.