gpt-5 is incredible at coding, and really shines with the right prompting style:

Introducing Kitten TTS, a SOTA tiny text-to-speech model - Just 15M parameters - Runs without a GPU - Model size less than 25 MB - Multiple high-quality voices - Ultra-fast - even runs on low-end edge devices Github and HF links below

This looks like a really cool alternative to the official MCP Inspector: https://t.co/osqocvGMAT Supports a Chat tab (with local model support!). Good for testing out different models with your MCP server. Going to think about using this for my workshops 👀

To safeguard privacy, if the “Pin to Drive” feature is enabled and “Home” or “Work” is selected, should the car prompt for PINs before revealing the destination? @Tesla @elonmusk

We’re going to make Baby Grok @xAI, an app dedicated to kid-friendly content

wow… Kimi K2 is terrifying it’s an open-source LLM that codes better than most paid models and can run agents without falling apart I pushed it hard... here’s what it built:

i've been using a Github deep research bot to find the best open source devs point it at any repo. o3: - finds all the top contributors - runs deep research on each profile - highlights the most hireable people - sends me a digest 15 minute build. free to use in thread

Demo/Source code:

Grok 4 Heavy is better than any model available at identifying issues in your codebase. Here's the JS prompt I use with my game code to have Grok 4 Heavy find the bugs. Python prompt in Comments👇

What cursor/claude/agent rules have made the most impact on your productivity when building with AI?

Web scraping will never be the same! Firecrawl is an open-source framework that takes a URL, crawls it, and converts it into a clean markdown or structured format. 100% Open Source

this is the value of MCP btw

🎉 Excited to announce that Atomic Testing now supports @vuejs alongside @reactjs ! Simplifying UI testing with a familiar API across frameworks, testing methods & design systems. More real-world examples dropping soon as we wrap up v1.0. Stay tuned! 🚀 🔗

Nice - my AI startup school talk is now up! Chapters: 0:00 Imo fair to say that software is changing quite fundamentally again. LLMs are a new kind of computer, and you program them *in English*. Hence I think they are well deserving of a major version upgrade in terms of

Browsers are the perfect place for hybrid AI inference, combining the power of cloud AI with the versatility and privacy of local models. ⚡️ Dia already has great WebGPU support, so I'm looking forward to seeing more on-device models being integrated directly into the browser

a=(x,y,d=mag(k=(4+sin(y*2-t)*3)*cos(x/29),e=y/8-13))=>point((q=3*sin(k*2)+.3/k+sin(y/25)*k*(9+4*sin(e*9-d*3+t*2)))+30*cos(c=d-t)+200,q*sin(c)+d*39-220) t=0,draw=$=>{t||createCanvas(w=400,w);background(9).stroke(w,96);for(t+=PI/240,i=1e4;i--;)a(i,i/235)}//#つぶやきProcessing

Introducing MCP UI, the SDK that wires UI components directly into the MCP protocol🤯 Watch the video to see the new phase in AI interaction! Why browse when you can have universal apps that deliver the right UI to consume any data and perform any action? Replace cumbersome

in R3F it's easy to construct these. there are helpers and various shapes (rect, circles, rings, etc). not only for testing, self made env maps are the key for good looking scenes bc it is essentially photographic lighting, or "light painting". see → https://t.co/Ks6f7o8F2d

https://t.co/TSSS8votZI

#mcp