Note: An LLM was not used in writing this microblog entry.
This entry is part of a microblog series called LLMs
Harvard published this article defining “workslop”, a new term for when someone substitutes hard thinking for LLM generated slop. I had been thinking about this recently, having witnessed it at work, but hadn’t articulated it yet, so I’m glad for the new term.
I accidentally discovered this article while looking up Holly Herndon, I enjoyed the reframing of LLMs from “artificial intelligence” to “collective intelligence”, which more accurately describes what a large language model is, and sounds a bit more positive and inclusive.
I got llama.cpp working from Emacs for both regular queries and GBNF-constrained outputs. I’m happy with the outputs, though I’ve not been using it for a couple weeks. Will come back to it.
Current mood: still interested to dabble more. Currently sceptical that organisations across the board are benefitting from LLMs. I read in the Economist that more small, boutique models, aka SLMs, are becoming preferred over the Big models provided by OpenAI, Anthropic, etc. which makes some sense.