Test Driving Code with the Help of AI
summary
Demonstrates using AI (VS Code/Copilot style) to write TDD for Conway’s Game of Life: starts with a canary test, has AI generate tests and methods like nextState, generateSignals, countSignals and nextGeneration, and implements them until tests pass. Shows useful patterns—select many/flatMap and group‑by for counting—and warns AI copies boilerplate, writes useless “not null” tests, and suffers a starting/stopping problem that requires active human steering.
AI-written summary · about
chapters (14)
- 0:00Test driving with AI
- 3:25Automated tests matter
- 7:39Canary test example
- 9:05Designing code incrementally
- 13:43Stateless design and AI
- 21:14First meaningful test
- 24:46Game of Life universe
- 26:12Generating live cell signals
- 33:16Functional style with AI
- 37:32Counting signals and tests
- 42:13Method interfaces via tests
- 48:56Next-gen Game of Life tests
- 55:01AI test correctness
- 56:07Summary
related talks
47mTest-driven Development, AI, Pair Programming, Agile Testing
26mAI and the Future of Testing
45mTest Driven Prompt Programming: How AI is Changing the Way We Code
01h 06mDoes GenAI Mean the Death of Test-Driven Development (TDD)
32mQuality engineering in the AI age: From testing AI to leveraging GenAI
38mWriting code live with AI
45mTDD with AI?
43mExploratory Testing with AI
