In their earliest iterations, Large Language Models (LLMs) didn’t have the exaggerated mannerisms of a helpful assistant that have since become their hallmark. They were limited to finishing sentences or curtly answering questions, greatly lowering their efficacy. This turned around with the development of “InstructGPT” in January 2022, the

The roots of this project come from my time working as a software engineer. I always found it bizarre how a few lines of code and a few pushes of a button from my apartment in San Francisco could affect real, concrete processes across the country. I suppose this is

Hi it’s been a couple of weeks since I set this up and things seem to have stabilized, so I thought I would write a short article on what exactly it is that I’m trying to do here and maybe kickstart the blog so people start writing more!