Zyphra have launched their first Text to Speech model, Zonos-v0.1 - now the leading open weights Text to Speech model in Artificial Analysis Speech Arena. Zonos-v0.1 is currently scoring a Speech Arena ELO of 1020, just behind proprietary speech models available on Google

grok 3 prompt: I’d like to make a p5.js simulation of a sphere made up of ASCII numbers, rotating. The closest numbers should be pure white, and the farthest ones should fade to gray, on a black background

Floods devastate communities around the globe, but Google Research is committed to leveraging AI to improve flood forecasting and response. Today, we shared that we’re expanding Flood Hub with new features: - Inundation History Maps - Basin View Per Gauge - Improved Layers Panel

creativity will scales like never before

post trained DeepSeek R1 with near-zero censorship curious about the fine tuning methods used to achieve this🧐

I think the question isn't if they can make $1M freelancing, but when

Claude 3.5 Sonnet still holding its ground after 8 months. oAI’s newest model can’t even dethrone it on SWE Lancer

cool event. now lets make sure this tech isnt locked behind paywalls or serving only corporate interests😅

we talk about AI alignment a lot, but maybe we should focus more on aligning AI research itself

⚠️ BRAIN ROT ALERT how large language models work, 100% explained: > transformer block > query, key, values > embeddings and positional encodings > multi-head attention > layer by layer operations

gm saturdays are for building

If this works at scale, AI governance is going to have a really bad time

Almost like they don’t want us to see how they really work🫢

GPT-4 looking at GPT-4.5 like a goldfish watching a blue whale

7th grader is out here deploying Grok 3 apps while I still struggle with npm dependencies🤣

trick question. You panic, say "uhh," and ask her favorite programming language

Can’t wait for my humanoid robot to take over my chores and then unionize against me😹

new brand is a vibe

ability to choose between fast and deep reasoning in AI is a gamechanger. The next step is to make it intuitive for humans too