Coding Tomorrow
summary
Claims concrete performance wins (Native image startup up to 100x faster), hands-on Java connectivity via LChain for J, and demos that actually run: an LLM returning a LocalDateTime object from messy text, Copilot/Cursor producing HTML+CSS+JS that renders, and an SQL generator that works most times but fails about 1 in 10. Built examples include a deployed Quarkus feedback analyzer and a couples-advice prototype; one gaming-PC hosts a coding model serving ~200 developers at ~100W. Warnings: hallucinations, higher code churn and imperfect PR reviews.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:08AI coding power demo
- 8:21How LLMs work
- 13:01AI in software lifecycle
- 23:16AI collaboration tools
- 31:00AI adoption challenges
- 34:10Java AI applications
- 41:59Future outlook
related talks
51mAI-Powered Development: Hands-On Techniques for Immediate Impact
52mLivecoding: Creating Beautiful Java Code With LLM and Agents
53mHigh Speed Java Development with LLMs and Agents Fast, Clean, Maintainable
04h 12mJava ON 2024 Day 2
42mCrafting Intelligent Applications in Java with Quarkus / LangChain4j
02h 49mJava meets AI: Build LLM-Powered Apps with LangChain4j
07mBest Practices and Key Takeaways
36mCrafting Intelligent Cloud Native applications with Quarkus