In early 2023 a paper was released describing recent advances in Large Language Models (sparks of AGI).  In summer 2023, I began experimenting with the models myself.  In addition to the web interfaces, I began installing open-source models on my local machine, and poking around with how they work.

LLMs are a very significant advance in AI.  These are not just stochastic parrots regurgitating words based on statistical probability.  The models used to generate those statistics represent an understanding of a world, and they can pull meaning from many different contexts.

I began to wonder what exactly could be accomplished using these tools as they continue to rapidly improve.