thingfyi

This is so cool! Definitely liked the title font :)

I do wonder if this kind of republishing is compliant with the original site? As in, if one can store the content vs. not and just lazily convert live - have been curious about the copyright rules & practices in this area.

This is a big finding.

Omar and his team has been working on this line of research (e.g., DSPy) for the past couple of years. To be honest, I had always been skeptical about his approach of using LLM to optimize LLM. It was in my view borrowing Matei Zaharia's research - Matei is Omar's PhD advisor, also the CTO of Databrick - in big data (compiler optimization etc.) and forcing it to LLM systems. Along with some other projects from James Zou at Stanford etc. in the same line of intuition.

Now, if this new result holds, I believe it's a huge deal for AI research. It means that LLM systems can truly runtime self-optimize/improve/adapt. Congrats to Omar and his team. It really takes year to punch it through and convince the doubters :)

> When looking at other large companies today, I noted that most of them are focused on selling to institutions rather than consumers.

This really shows that one that owns the consumer devices has a big leverage to play even in the age of AI - the entry to almost everything.

Very cool project! I also wonder if a web version is plausible. Are there any specific native app features in it?