Navigating AI assistant development
summary
Walks through using GitHub Copilot inside the editor to generate C# boilerplate—MockMVC setup, test data builder generation and Swagger/Testcontainers snippets—and shows where generated integration-test assertions, Flyway migrations and entity mappings go wrong. Demonstrates iterative prompting, keeping TDD, and the need for a human review loop to avoid hallucinations, context poisoning and automation bias. Cites a Thoughtworks survey of 50 live projects showing mixed productivity effects and recommends small, measured adoption.
AI-written summary · about
chapters (8)
- 0:00Generative AI overview
- 4:02AI coding assistants
- 10:38GitHub Copilot demo
- 15:17Code quality evaluation
- 21:13Engineering in AI era
- 24:43Productivity impact
- 30:01Team adoption survey
- 31:46Key takeaways & ethics
related talks
38mNavigating AI assistant development
01h 02mSkills for the age of AI developer tools
36mME, MYSELF AND AI
25mImproving developer happiness with AI
33mEnsuring quality in an AI-driven world
28mBoost your development workflow with AI
14mDeveloper productivity in a post-AI world
26mWhen Agile and AI Collide