Thought of checking Bsky again… any good poasters I should follow?

Elon advocating for a 6mo pause in training new AI models

can we come up with a way to do distributed training of models on hardware contributed in a similar manner to mining pools? Aware we need a different paradigm for gradient descent… so what would that look like?

How good’s your unicorn, anon?

This feels like leaving my footprints on the moon - haven’t joined a “new” social network in 15years lol