AI and Code Quality: Building a Synergy with Human Intelligence
summary
The talk examines how AI (using OpenAI models integrated into VS Code via custom plugins and Copilot) can analyze recent repository code to extract and propose team-specific coding practices and detect issues (examples: enforcing function names that reflect side effects, flagging tests that exercise multiple methods). It demonstrates an IDE plugin that captures such examples and reports AI-detected problems, while warning that AI output quality depends on input data and team standards, detections are non-deterministic and need human validation, and productivity gains vary (Copilot can speed individual tasks by over 50% but studies show only about 10% overall productivity improvement).
AI-written summary · about
chapters (11)
- 0:00AI and code introduction
- 3:51AI revolution in dev
- 8:06Technical debt risks
- 13:23Real productivity impact
- 17:33Global vs local optimization
- 21:21Capturing best practices
- 25:10Tacit knowledge sharing
- 33:10AI as teaching tool
- 37:22AI use cases
- 40:10Human expertise value
- 44:02Q&A
related talks
34mAI and Code Quality: Building a Synergy with Human Intelligence
25mFrom AI Chaos to Production Ready: Using AI Agents to Improve Your Code Quality
46mAI Integration for Devs: ChatGPT and Copilot in Your Projects
14mDeveloper productivity in a post-AI world
39mPower of AI and Code Quality and Performance
25mImproving developer happiness with AI
15mAI's Impact on Development, Post-Production & Support Journeys
19mWith great power comes great responsibility in code quality