Erik Voorhees (@ErikVoorhees) started following @vincentweisser on Oct 8, 2024

RT @Grimezsz: @PrimeIntellect Omg haha - we're just the sequel

RT @vincentweisser: You can just start a hero run with 256 H100s on-demand on https://t.co/zgGDdtG9Rz https://t.co/TYYNgnYpVv

Friends, don't let what happened to #NorthCarolina happen to the greater #TampaBay area. It's very early, but the models show a potential worse case scenario for the #Florida W Coast. Let's stay focused, be prepared, and don't let the #NorthCarolinaFlood fall off the headlines! https://t.co/G0uk4vEAzX

Stop us? Mike says NO! #CajunNavy https://t.co/W68XnzX0gY https://t.co/CgfTNALNcX

Please read this important and poetic essay written by someone who actually understands both AI and bio

And the two years after that will happen even faster. And so on, and so on. That’s what it’s like to live in accelerating times

RT @karinanguyen_: My vision for the ultimate AGI interface is a blank canvas. The one that evolves, self-morphs over time with human prefe…

RT @rauchg: Example of a @vercel customer now saving 35% of their Functions bill after enabling In-function Concurrency 😳 https://t.co/1s01…

De graça leia e tire suas conclusões

I hurt my shoulder 3 weeks ago… but I couldn’t pull out of a main event… thank you my team for helping me training like that e get me ready for the fight #ufcparis #att

*technically I invested in Chronos Research and the consumer product that would become Robinhood was called "Cash Cats". Vlad and Baiju wouldn't come up with the name or even complete idea for Robinhood for another year.

I was the first investor in @RobinhoodApp back in 2012. Excited to back @vladtenev again with @HarmonicMath Whether you call it Q* or reward function, accurate feedback for model training is the key missing ingredient to get to AGI. Harmonic is developing advanced math models

RT @SportsCenter: Chicago Bulls Derrick Rose was different 😤 https://t.co/zjaVrh9QwJ

RT @elonmusk: How time flies! https://t.co/xgl7Zr05tR

Evaluating performance on extensive human evaluations and benchmarks, results suggest that Llama 3.2 vision models are competitive with leading closed models on image recognition + a range of visual understanding tasks. https://t.co/QtOzExBcrd

This is just a small slice of what’s new with our Llama vision models. You can find more details in the full model card published on GitHub ⬇️ https://t.co/agWQxRBLP9

RT @ST0NEHENGE: Stonehenge Autumn Equinox https://t.co/A5ouvQQgpN

RT @ST0NEHENGE: Autumn Equinox Stonehenge : https://t.co/mXqG0OCaoB

RT @TedMerz: View from the Office. I met Craig DeWitt, co-founder of startup Skyfire, for breakfast at Balthazar. I had the omelette with…

We’re adding more services and providers weekly. If you’re building valuable AI services and want to monetize your API Or if you want on demand access to premium agents, DM!

RT @0x_pix: i don’t need to spend $90 on this. i don’t need to spend $90 on this. i don’t need to spend $90 on this. https://t.co/7eHWhfM…

RT @vxnuaj: Saw many great things at NousCon. Good stuff @NousResearch. DisTrO gonna go crazy. https://t.co/8ucp6BWTOS

I have discovered a truly marvelous proof of AGI, which this Hugging Face model repo is too narrow to contain...

If anyone has been wondering what I've been up to for the past six months, the answer is Hermes 3 and DisTrO 🤗 Really excited to actually be able to show this off and start talking about it

RT @rivatez: had a personal emergency last night and couldn’t give my talk at NousCon. thank you so much to the amazing @apyh__ for deliver…

RT @haizelabs: We're excited to share our new preprint introducing endless jailbreaks via bijection learning. Our attack exploits the adv…