My 6885 Bookmarks & Likes

META JUST KILLED TOKENIZATION !!! A few hours ago they released "Byte Latent Transformer". A tokenizer free architecture that dynamically encodes Bytes into Patches and achieves better inference efficiency and robustness! (I was just talking about how we need dynamic… https://t.co/aawPMIRjjG

In one year, I genetically engineered the boring Arabidopsis into a beautiful ornamental flower 🧬🌹 Here is how and why I gave this model organism a visual upgrade 🧵(1/7) https://t.co/Od3ysI9gqM

I work for the government. As a technologist, I cannot keep this under wraps anymore. There is a government-funded organization that is already experimenting with a quantum computer they call Miya, and it is orders of magnitude more powerful than Google's Willow. https://t.co/84dtVW5N1G

What could the future of human-agent interaction look like in your browser? 🌐 Project Mariner is a research prototype built with Gemini 2.0 that’s able to research information and carry out tasks directed by you through an experimental Chrome extension. See it in action ↓ https://t.co/HkJ54hOpxk

Can anyone explain why? Is it related to Willow the quantum chip? Thx I'm noob https://t.co/cdukEWY7Vd

Introducing Willow, our new state-of-the-art quantum computing chip with a breakthrough that can reduce errors exponentially as we scale up using more qubits, cracking a 30-year challenge in the field. In benchmark tests, Willow solved a standard computation in <5 mins that would…

I don't know if people realize how literally I mean it when I say that AI agents are going to turn building a business into playing a game of Factorio https://t.co/r0tsNie6ry

Your first programming language shapes they way you solve problems. Really interesting read. https://t.co/JB189lIRQj

As we continue to explore new post-training techniques, today we're releasing Llama 3.3 — a new open source model that delivers leading performance and quality across text-based use cases such as synthetic data generation at a fraction of the inference cost. https://t.co/BNoV2czGKL

I've made projects with: - health stats so Claude can be my doctor - financial info so Claude can advise me - goals/resolutions so Claude can coach me - research papers so Claude can teach me - books so Claude can mentor me Claude with Projects is a literal life hack

Dr Louks, we support you. Last week, Cambridge PhD student @DrAllyLouks published this photo on X to celebrate passing her viva with no corrections. Her tweet went viral, attracting over 100 million views 🧵 https://t.co/W2gPHSY1yl

I think most people don't understand how huge this is. Let me try to explain. We have compressed most human knowledge into a single model. They have some limitations, like they aren't the best at math and planning yet, but bear with me. It all boils down to emergence. With a 26… https://t.co/AerQlAiCWS

hey bro. not a big deal but just checking in because it seems like your arc is stuck on filler episodes. i mean they’re cool and it’s nothing urgent but just wondering when the plot is going to start advancing again

underrated repo, good variety of papers https://t.co/KaV1QgzkKm

Curiosity requires slack in your life Don’t become too employed

Introducing DeepThought-8B: Transparent reasoning model built on LLaMA-3.1 with test-time compute scaling. - JSON-structured thought chains & controllable inference paths. - ~16GB VRAM, competitive w/ 70B models. - Open model weights, and inference scripts. https://t.co/IMHnKqJc2x

It’s done! 150,000 words, 200+ illustrations, 250 footnotes, and over 1200 reference links. My editor just told me the manuscript has been sent to the printers. - The ebook will be coming out later this week. - Paperback copies should be available in a few weeks (hopefully… https://t.co/i8g2jvRc0D

Introducing Fish Speech 1.5 🎉 - Making state-of-the-art TTS accessible to everyone! Highlights: - #2 ranked on TTS-Arena (as "Anonymous Sparkle") - 1M hours of multilingual training data - 13 languages supported, including English, Chinese, Japanese & more - <150ms latency with… https://t.co/Oz9BNHBPEf

Introducing Genie 2: our AI model that can create an endless variety of playable 3D worlds - all from a single image. 🖼️ These types of large-scale foundation world models could enable future agents to be trained and evaluated in an endless number of virtual environments. →… https://t.co/qHCT6jqb1W

Today in @Nature, we’re presenting GenCast: our new AI weather model which gives us the probabilities of different weather conditions up to 15 days ahead with state-of-the-art accuracy. ☁️⚡ Here’s how the technology works. 🧵https://t.co/PWCNWbQnlU https://t.co/6DTrmn64Jq

the universally appreciated design vibe right now seems to be “retro future” and “future retro” — described closest by the concept of “anemoia” https://t.co/2KPEJtBqzg

Starlink + robotics + weather forecasting = autonomous wind-powered cargo vessels. Zero emission maritime shipping, faster and cheaper than what's possible now. Say hello to Clippership. https://t.co/2ayjQf4QZf

Announcing Exa Websets - a breakthrough toward perfect web search. Sign up for the waitlist below👇 https://t.co/BHK97QX4Sc

Conversational AI is here. Build AI agents that can speak in minutes with low latency, full configurability, and seamless scalability. https://t.co/JqBlwVczdX

Just do it. https://t.co/lvKNYxz2oW

India strikes $715m deal with 30 publishers -- including Elsevier, Springer Nature and Wiley -- giving 18 million academics nationwide free access to 13,000 research journals from 2025. https://t.co/I8RI416PI3

After a ton of research, here's the initial Operate stack (some still tbd) ⛺️ @supabase - db 🛜 @vercel - host, img, analytics + next 📨 @loops - emails 🤖 @inngest - bg jobs ⚠️ @knocklabs - notifications ⚡️ @powersync_ - sync engine 🪲 @jamdotdev - bugs 📊 @IndexBI - bi

Starting an online business today is insanely cheap: • Figma: $0 • Next.js: $0 • Supabase: $0 (up to 50k users) • Resend: $0 (up to 3k emails/month) • Domain: $10 • Stripe: $0 (1.5%-2.5% fee) For $10 and a few hours a day, you’re one step away from building something…

X Article: https://t.co/CXiOq99RuW

"just do it" https://t.co/jzxpWI4xwN