Unlocks Calendar

@LongzuAlpha 📸

@AndrewCSAI fuck u

This is a game-changer announcement by Apple around cryptography. It is the “HTTPS moment for AI” in some ways.. Here is what this means: your private confidential data can be pooled with other data sources and used to securely improve your UX and that of the wider community without even revealing your private data at all: it stays encrypted end-to-end. This is a win-win, a rare non-zero-sum game in technology. For example, consider a picture on your device. It gets converted to vector embeddings (the numbers you see below), then encrypted and sent to the “Homomorphic Encryption” (HE) process running on a server which returns a fully encrypted result back, without decrypting (revealing) the encrypted user private data sent initially - even for it’s own processing. This is search over fully encrypted data! The user gets a notification that the picture region here looks like the Eiffel Tower in this example. Thus this system enables using the network hivemind to securely to improve the UX across so many applications in the time ahead. It will especially benefit AI products which have a direct relationship with end-users as now the standalone tool with access to confidential data can also benefit from network effects for optimal UX. Information wants to be free and mashed together endlessly, and this pushes in that direction. cc @AravSrinivas @sama -> FHE is a key enabling technology over a 5 year horizon. This is how you leapfrog Google.

Just got o1-prev access (many thanks to @AlexanderMoini). Here are the results! 1. Massive library 2. Floating city 3. Burj Khalifa 4. Literally prompt "Something cool" The model is very prompt adherent. I would say better than sonnet 3.6, but much slower and less creative.

Time for an American to respond: iMessage: 2 minutes Slack: 10 minutes Email: 12 hours WhatsApp: 7-10 business days

Dreamina AI V2.0: A powerful, all-in-one AI generator Developed by ByteDance, the parent company of TikTok, Dreamina AI is currently in beta testing for its new version 2.0. Check out their 1.0 link below. This updated version offers a wide range of features, including image and video generation, AI music creation, storyboarding, motion brush, canvas tools, and more. Essentially covering all the major creative functions you’d expect from AI. Link below

🚀 Excited to announce that DeSTA2, a spoken LLM. What's amazing? It's trained on just 150 hours of speech data, beating models with tens of thousands of hours! Shoutout to NTU PhD student Ke-Han Lu & NVIDIA team for leading this! 👉 Learn more: kehanlu.github.io/DeSTA2/

We started our 10b training Decentralized. We only need one minute of communication every hour of training. Locked in for 3 weeks, long night of no sleep but it's finally live

GPT-NeoX is already the go-to library for pretraining at scale for dozens of organizations globally. Now with @synth_labs we're adding native RLHF support to give you even more reasons to try it! We find GPT-NeoX is 30% faster than trl at the 13B scale. blog.eleuther.ai/rlhf-and-rlaif…

Artifacts, but 100% Open Source, Private & Local. @OpenWebUI has released a game changer update---"Artifacts". Instead of proprietary LLMs, you can now use fully LOCAL and PRIVATE LLM via @ollama to build apps! Here I use @Alibaba_Qwen 2.5 32B to build a game of life app.

My 7 yo and I spent 25 minutes building a tic-tac-toe app with AI. He called it "the most interesting thing I've done in my life." It has easy/hard mode, happy confetti, a scoreboard, and sound effects. And we recorded it all live. If he can do it, what's your excuse?

Chinese startup Leju Robotics has released their open-source humanoid development platform for academic and R&D use cases. It includes an SDK for sensors and controls, simulation models, an LLM interface, and some basic demos that work out-of-the-box.

An AI will win a Nobel price someday✨. Yet currently, alignment reduces creativity. Our new @GoogleDeepMind paper "diversity-rewarded CFG distillation" improves quality AND diversity for music, via distillation of test-time compute, RL with a diversity reward, and model merging. arxiv: arxiv.org/abs/2410.06084 website: google-research.github.io/seanet/musiclm…

How your dreams feel after spending all day on CT

openai’s cost structure

Training Llama 3.1 on clinician-created synthetic data, using prompt engineering techniques and RAG; Neuromnia developed Nia: a human-centric AI co-pilot to support work on some of the most pressing challenges for autism care ➡️ go.fb.me/ow8n1d

Three Lobster Moon by Ilya Sutskever

Legendary NYT interview by Geoffrey Hinton

i literally cried. i’m so happy for him. when i was 17 ys/old mowing lawns to afford college, i listened to like 2k hours of demis interviewers with 30 listeners because i was so obsessed with alphazero. i hope he realizes how much beauty he’s brought to science