madness and machines

Wintertime Reflections on AI

For all of the opaqueness surrounding them, for all the prognostication of AI boosters who predict various kinds of societal

I trained my first language model in 2016. Back then, we didn't distinguish between small language models (SLMs)