The Era of AAP: Ai Augmented Programming using only Java
summary
Demo of a Java-first AAP stack: an open-source IntelliJ plugin Devox Genie that runs local Llama models or routes to cloud providers, nearly 4,000 downloads, inspects a 149-file project (990k tokens) and generates a README in ~12s. Java inference projects JLlama and llama3-java perform pure-Java model inference, support sharding (405B model split across machines) and aim for ~20 tokens/sec streaming. Notes on Panama/FFM, TornadoVM GPU work, 2M‑token contexts (Google Gemini) and a ~€15 cost per 2M‑token request are included.
AI-written summary · about
chapters (7)
- 0:00AI Augmented Programming intro
- 4:16AI coding and privacy
- 10:56LLM context windows evolution
- 18:25Java AI coding systems
- 25:51LLM advancements opportunities
- 33:01Java LLM IDE integration
- 37:19Q&A
related talks
46mThe Era of AAP: Ai Augmented Programming using only Java
46mThe Era of AAP: Ai Augmented Programming using Modern Java
46mThe Era of AAP: Ai Augmented Programming using only Java
58mGenesis Keynote
47mJVM in the Age of AI: Babylon, Valhalla, TornadoVM and friends
47mPractical LLM Inference in Modern Java
50mGenerative AI in Practice: Concrete LLM Use Cases in Java with the PaLM API
30mDevoxxGenie: Your AI Assistant for IDEA