Aella
RT @JeffLadish: I don't necessarily like the changes themselves, but I am happy to see Elon experimenting with Twitter features

Try stuff, keep what works and scrap what doesn't work. Twitter was too conservative before
tweet

@WeavingWithAI@sigmoid.social
RT @uwepleban: @NathanLands @GanWeaving @LoreAIArt Technological utopia, V1 through V4. https://t.co/15I7vQYFH2
tweet

Mohammad Bavarian
A core component of Effective Altruism doctrine is "earning to give".

But while "earning to give", ensure you also are "giving to earn". That is, make sure you earn by actually providing real value to the people. If SBF had followed this, the disasters would've been avoided.
tweet

Bojan Tunguz
If you are curious to find out what my Ph.D. Thesis was all about, need a last minute Christmas gift, or are suffering from a chronic insomnia, you can order your very own paperback copy from Amazon:

https://t.co/0tMpARfG1t https://t.co/ZCHvTVPUWG
tweet

trash
RT @mileswjohnson: Are @trashh_dev fans called trash cans? https://t.co/ZWgeceo4bO
tweet

Hugging Face
RT @yvrjsharma: Exciting news! I'm thrilled to announce that I'm now an ML Engineer and Developer Advocate at @huggingface, having join the @Gradio team!
I've only been here for a week,but I'm already loving the team and the projects we're working on. Can't wait to see what the future holds🤗🔥
tweet

gi
nothing is as it seems because everything behaves differently under observation
tweet

Stable Diffusion 🎨 AI Art Pics
Wearing gold armor is a lousy idea, but it certainly looks cool. #AIArt

#StableDiffusion2 / #StableDiffusion #DreamStudio https://t.co/ZLadiwZs2F
tweet

Deen Kun A.
FAIR recently released a version of OPT instruction fine-tuned on ~2K tasks. The paper is more interesting than model, IMO. Highlighted what worked well & what didn't.

In it, is an interesting table showing 11B FlanT5 outperformance.

NB: in MMLU, code-davinci exceeds Flan-PALM https://t.co/UFwxFKkN19
tweet