Copilot in Android: AI for boring stuff, people - for important
summary
Demonstrates using GitHub Copilot to scaffold an OpenWeather API app: it auto‑writes Gradle files and dependencies, generates Hilt modules, Retrofit interfaces, data classes and mappers, utility functions and a validating regex, cutting boilerplate setup to minutes. Claims the model trained on 700 million lines of public code, works from current/open files only, needs internet, and often produces plausible but buggy snippets that require manual review; useful for routine tasks but not architecture.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker intro
- 2:56What is GitHub Copilot
- 9:01Setting up new projects
- 14:19Best use cases for Copilot
- 21:05Improving Copilot outputs
- 23:37Limitations and issues
- 29:22AI ethics and intellectual property
- 31:33Conclusion and takeaways
related talks
38mNavigating AI assistant development
50mHow I Made Peace with AI
34mAI won't take your job but it sure will help you do it better
55mCopilot is my Copilot: Pair programming with an AI
57mPut an Agent Inside Your App in 10 Minutes or Less with the GitHub Copilot SDK
42mBuilding an AI companion with Modern Android Architecture
01h 00mGitHub Copilot, using AI to help you learn, code, and build
25mImproving developer happiness with AI