AI can generate code at unprecedented speed, but it can also generate bugs and regressions at unprecedented speed. As software creation becomes cheaper, the bottleneck shifts from writing code to verifying behavior.
This talk revisits the rise and fall of Cucumber and BDD, exploring why executable specifications struggled to gain adoption and how AI may have changed the economics behind them. We’ll discuss how Ruby teams can build practical testing pipelines for AI-generated code using AI-assisted system tests as a safety net.
Cristian Planas is a Ruby developer, conference speaker, and author of *Rails Scales!*. With over 15 years of Rails experience, he has helped build and scale large applications at Zendesk and now works at Combinaut, exploring the intersection of Ruby, testing, and AI-assisted development.