God is always with us .

China's Bytedance just dropped an AI video editor that understands video better than even Gemini 3 Pro. Vidi2 can take in a bunch of footage many hours long and a prompt, and construct a script and generate a TikTok or movie from them.

Vidi2: A 12B multimodal model for video understanding & creation; up to 30 mins; enables plot-aware auto-editing. - Automated "Long-to-Short" Repurposing - Smart Cropping New level for tiktokers. bytedance.github.io/vidi-website/

ToolOrchestra is such a cool work from NVIDIA Just an 8B model trained on calling tools and other LLMs to answer queries It's a great demo of what frontier SLMs will be about in 2026

Replace assets right inside your block with just a click. No more dropping new images and connecting new noodles all over. Right click and select "Replace Block" or hit command + shift + R.

To everyone wondering why there are no AI movies or TV shows yet; the economics and the technology don’t yet support it. The tech evolves so fast that you’d constantly have to reshoot your film. We’ve tried this and it’s 4+ months and $300k+ to make a 30 minute AI episode that

LTX-2 is like Z-image for the video community: fast and exciting. Anyway, there are some improvements; extracted parts from the checkpoint: - ltx-2-19b-dev-fp4 video VAE - vocoder - projections to use the native ComfyUI DualClipLoader with gemma_3_12B. huggingface.co/GitMylo/LTX-2-…

Islamic intolerants know that they have always been at war with the culture(s) of Iran which Iranians and Persians carry with them in the diaspora. They massacred 11 souls here in this one famous cosmopolitan cafe in Colaba. Just massacred them. Tourists. Indians. Persians.

We just published the largest review of creatine for brain health. Increased brain creatine levels may function as a buffer to metabolic stress leading to neurological benefits. Doses >5g/day are likely required for brain benefits.

Using Claude Code with a three.js skill With proper 3D web context, Claude reasons in terms of scenes, interactions, and WebGL constraints instead of isolated snippets. Install it with: npx claude-code-templates@latest --skill=creative-design/3d-web-experience --yes More

👀

Reverse engineering the binary just to find a hidden websocket flag is the most based thing i’ve seen this week. @anthropic tries to gatekeep the ui but the infra always finds a way. reminds me of why we used zmq for sglang—sometimes the simplest transport layer is the most

pi-coding-agent from Mario Zechner is magic. just tell it what functionality you want it to extend itself with and it will. can imagine a future where every programmer carefully curates their pi setup just like they always have for .vimrc published a small extension, pIRS,

Wow, I guess 1.1 million of you found this useful. Thank you. The age of the Code Factory is here.

Content Artist, Blake Johnson, created this epic bridge scene using Luma Agents, blending it with his traditional 3D workflow to bring the shot to life. Behind-the-scenes coming next! Luma

This is what directing an AI creative studio looks like. I gave my orchestration system a reference image and had it deploy three different interpretations across three machines running Blender simultaneously, all visible from a single dashboard. When one direction looks

This is actually insane. Dude hard-coded a WebAssembly (WASM) interpreter into the weights of a transformer, losslessly. In essence, a computer is running inside a LLM that can actually run computations, not infer or guess a calculation like most do today.

Oh wow. Defenseless against savage Islamists? “Can’t imagine how that feels,” said none of the 90,000,000 Iranians who have endured 47 years of this repression.

LGTM from Apple: 4K feed-forward 3D Gaussian Splatting. instant 4K 3D scenes without massive GPUs.. - predicts a few lightweight 3D shapes, wraps them in ultra-high-res 2D textures. - low Memory usage You take two normal photos of a room. Instantly walk around it in flawless