Lesson 19: Verification, Evals and Knowing When the Job Is Finished
Two AI-generated solutions can look equally impressive: both have tidy code, polished screens and confident explanations. Open Full Article...
Home schooling resources for children and parents.
Two AI-generated solutions can look equally impressive: both have tidy code, polished screens and confident explanations. Open Full Article...
This guide explains why AI-generated code needs careful review before acceptance. It distinguishes a working result from sufficient evidence for approval. Open Full Article...
This guide supports teaching learners to review AI-generated code and the behaviour of the agent that produced it. Open Full Article...
An AI coding agent reports, “Feature complete: all tests pass,” yet its diff adds an unrelated dependency, changes a public function, weakens error… Open Full Article...
This lesson explains why AI-generated code is an implementation that still needs verification. Open Full Article...
This guide helps you teach why AI-generated code is a hypothesis until relevant evidence shows that it meets its requirements. Open Full Article...
An AI coding agent confidently announces, “This should work,” just as a previously reliable feature quietly stops working. Open Full Article...
This guide explains why changing existing software requires care. It defines refactoring and distinguishes it from feature development and bug fixing. Open Full Article...
This guide supports teaching learners aged 11 to 16 how to judge AI suggestions for changes to existing software. Open Full Article...
Imagine an AI rewriting a difficult function while also reorganising directories, renaming public methods, changing a dependency and altering behaviour. Open Full Article...