Register - Login - Forgotten Password
by

Lesson 15: Testing, Verification and Making AI Prove Its Work

“Finished,” announces an AI coding agent, but the build has failed, the feature behaves differently in one browser and one acceptance criterion was never checked. How can software be complete if the evidence says otherwise? This lesson distinguishes an implementation—the code intended to produce behaviour—from verified software, whose intended behaviour is supported by relevant evidence. The central question is simple: what proves that an AI-generated feature really meets its requirements?

Lesson 15: Testing, Verification and Making AI Prove Its Work

This lesson just one in the category Vibe Coding and AI-Assisted Software Development. To access this lesson you need to have a valid learner account, ask your parent to register you now.