Learnings

My notes on generative AI concepts across languages, starting with Java and Spring AI, with room for examples in TypeScript, Python, and Go.

How I reduced my Docker images by rethinking layers, separating build and runtime stages, and using Next.js standalone output and Bun server builds.

How I shared React Native components with Astro and explored Next.js, including Uniwind styling, bundler differences, and the limits of my setup.

How I think about LLM hallucination: token prediction, missing evidence, ambiguous prompts, and why better context helps without guaranteeing a correct answer.

Why closures can retain memory after clearInterval, how garbage collection follows references, and how I approach cleanup and heap snapshots in JavaScript.