Foojay Podcast #74: JCON Report, Part 3 - AI, ChatGPT, LLM, ML, RAG, MCP, GenAI, and more
AI talk often stays abstract, but Java developers want to know which frameworks, protocols, and patterns actually hold up in real projects. At JCON in May, we cornered speakers between sessions to ask exactly that. In this third and final JCON recap, Pasha Finkelshteyn, Simone de Gijt, Steve Poole, Sandra Ahlgrimm, Mary Grygleski, Jonathan Vila, Simon Martinelli, and Emily Jiang join me for Foojay Podcast #74, Season 4.
What we talked about
- RAG (Retrieval-Augmented Generation) in Java projects
- MCP (Model Context Protocol) and what it changes for tooling
- Working with Large Language Models from the JVM
- Spring AI and LangChain4J as practical entry points
- Risks, dangers, and honest limits of current AI tools
- Infrastructure as Code in an AI-driven workflow
- How chat interfaces reshape UI development
- Java compared to Python for AI work
What stood out
The guests agree that Java has a real seat at the AI table, not just as a client to Python services. Frameworks like Spring AI and LangChain4J make RAG and MCP approachable for teams already on the JVM. The conversations also keep a healthy edge, naming the failure modes and trust issues that come with these tools.
See the Foojay Podcast #74 for all info, shownotes, links, etc.