Two recommendations for you.
Here’s Terence Tao with a great explanation of how LLMs work and the dangers (and benefits) of letting machines do our donkey work:
… and here’s Richard Campbell with some interesting takes on circular AI <-> compute investments, limits to future LLM improvements, and some things to be hopeful about:
See also:
Eye on AI: Platonic banana trouble
Eye on AI: Shy Kids, airhead slot machine
Eye on AI: Benedict Evans on automation and the Jevons paradox
Eye on AI: Ross Anderson on model collapse; or, The Curse of Recursion
Eye on AI: Ice Ice Matrix; ABBAtars
Eye on AI: ChatGPT and Me
Ted Chiang: AI ain’t art
Unreal City: T. S. Eliot’s Wasteland Jukebox feat. Dall-E [known to be the wisest woman in Europe]
The AI menace that no-one talks about
dr.ai.verless crocod.ai.l // Hype- Text Transfer Protocol