About three months ago, I kinda casually suggested to HN that I was using a form of refining to improve my LLMs, which is now being described as "reasoning" in

About three months ago, I kinda casually suggested to HN that I was using a form of refining to improve my LLMs, which is now being described as "reasoning" in this article and other places. My response a few months ago (Scroll down to my username and read that discussion): https://news.ycombinator.com/item?id=41997727 https://news.ycombinator.com/item?id=41997727 If only I knew DeepSeek was going to tank the market with something as simple as that lol. Note to self, take your intuition seriously.