gi
annoyed
tweet
Offshore
Video
Claire Silver 🌸
RT @MalcolmRenoldz: @greg_price11 @MostlyPeacefull It’s even worse than you think. https://t.co/xFElWOUVfd
tweet
Lior
AI search platform Perplexity just raised $25.6M for their series A!

"Our long-term mission is to become the best platform for answers and information, serving as the go-to source for people seeking quick, accurate answers tailored to their asks." https://t.co/wrx2pDk4JO
Excited to announce that we’ve raised 25.6M$ at a 150M$ valuation led by New Enterprise Associates, along with our seed round lead investors Elad Gil and Nat Friedman. Also honored to have @paultoo and @soleio partner in this round to help us build a https://t.co/5QlDbpjIvb… https://t.co/1GqHYs6sOD - Aravind Srinivas
tweet
Offshore
Photo
Claire Silver 🌸
RT @greg_price11: The bill to ban TikTok is absolutely terrifying. It gives the government the ability to go after anyone they deem as a national security risk at which point they can access everything from their computer to video games to their ring light.

This is a Patriot Act for the internet. https://t.co/uYea49F1b1
tweet
Offshore
Photo
UnknownCollector
RT @anonimlesmek: the artists buying their own art from 3 wallets be like https://t.co/bO8vPDUjgv
tweet
Jeremy Howard
fastai is now compatible with PyTorch 2.0, and its compiler, for ~50% speed improvements! https://t.co/B6zoSG80g5
New #fastai release: v2.7.12. https://t.co/97fKzjPHeP

New Features

- PyTorch 2.0 compatibility, thanks to @benjamin_warner
- Pytorch 2.0 compiler compatibility, thanks to ggosline
- Better version support for `https://t.co/S58jNand3A_empty`, thanks to @benj…
- fast.ai
tweet
Offshore
Photo
UnknownCollector
RT @grebenshi: waiting for godot https://t.co/NABbkRPd75
tweet
gi
i don't think love is supposed to feel like walking into a trap
tweet
Offshore
Photo
Jeremy Howard
"Optimal constrained by memory use" seems like the thing that everyone should be striving for, but very few are. https://t.co/329bVHPmoN
Recently I’ve been harping on how “compute optimal” and “best for use” are completely different things. This plot from @CerebrasSystems shows that really well: their compute optimal models trained on the Pile out-preform Pythia for fixed compute but underperform for fixed params https://t.co/NH14Vux9Tr - Stella Rose Biderman
tweet
Ferruccio Balestreri
RT @karrisaarinen: I think partly why the data driven development is to prevalent is because it's simple. It's easy to ask someone to deliver x% increase in y%.

It's a harder leadership challenge set a vision and ask for high quality. There is not great objective measure, it's mostly subjective.… https://t.co/tMzZWn14Sp
tweet