Increasing code habitability by using AI coding assistants
summary
Claims that AI coding assistants can improve habitability by concrete edits and tests using GitHub Copilot are demonstrated in a Ruby URL shortener demo where dependency injection replaces brittle globals, tests are rewritten for isolation, and an added collision test forces a robust retry fix so tests go green after fixes. Backing stats include 4.5 billion contributions, 988 million new projects and 74% developer favorability, and the demo shows assistants help but require careful incremental prompts and human judgment.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 1:03Higher bar for code
- 4:36Code habitability basics
- 8:27Demo introduction and codebase overview
- 17:20Demo recap
- 19:25AI assistant guidelines
- 21:11Closing
related talks
33mAI assisted software delivery
14mDeveloper productivity in a post-AI world
25mImproving developer happiness with AI
15mAI's Impact on Development, Post-Production & Support Journeys
51mAI-Assisted Coding: The Future; Between Challenges And Benefits
45mThe AI Revolution in Software Development: From Code Writing to Quality Assurance
45mAI Assistance for Software Teams: The State of Play
18mCoding Reinvented: How AI Agents Are Shaping the Future of Software Development