ASK HN: GO DEEP INTO AI/LLMS OR JUST USE THEM AS TOOLS? (HACKER NEWS THREAD)

Summary from the TLDR newsletter

Large language models are becoming commodities. Many people use them, but few people are developing them from scratch. Developing large language models involves a lot of hardcore computer science and math - people can use them without having to know any of the internals. While it might be helpful to understand some of the internals and math, it's probably best to stick to learning how to get good at using these models and integrating what they do into agentic systems.