You can one-shot a documentary on any topic using the @runwayml MCP:
"Make a five-minute 16:9 Netflix-style documentary explaining Runway's world models for beginners, hosted by Cris and Anastasis, interspersed with 3Blue1Brown-style animations."
i'm still not sure people are fully getting how good opus 5.5
i gave my claude agent fig access to the runway MCP and told it to make a 'high-end netflix style documentary about superintelligence for normies' and...
it came back with this
Claude can rap! He can even trap (see the last one)
I call it Tokenwave.
"Make a hip-hop music video using ffmpeg and python, and rap with say/espeak."
Claude can rap! He can even trap (see the last one)
I call it Tokenwave.
"Make a hip-hop music video using ffmpeg and python, and rap with say/espeak."
The Runway Model Router automatically chooses the best model for each generation, based on rules you set. Optimize for cost, quality or latency, cap what a single generation can spend, and skip model evaluation.
Here's how to set up your first Model Router.
Following last night’s opening in New York, ./spiral is now open online to everyone.
Explore new work by 11 artists across AI, generative and protocol-based art, and collect the remaining available works through SuperRare.
https://t.co/2t6ax5bopx 🌀
./spiral opens tonight in NYC.
Opening reception: 6 PM at Fondazione Aversano, 58 East 11th Street, ground floor.
The online release opens tomorrow at https://t.co/2t6ax5bopx.
A return, not a retrospective.
Explore the works below ↓
@genekogan presents Player Computer, 32 unique 1/1 videos inspired by the player piano. Code draws every frame and synthesizes every note, then travels with the film so viewers can fork it and cut their own “roll.” Kogan’s long-running subject is creative agency.
This is not just another video simulation with a WASD controller. The big new idea here is the WorldPrompt, which lets you program a live stream of text actions addressed to specific subjects or to the scene itself as it evolves, e.g. one player controls the protagonist, another the camera, another the weather or environment, etc. The interaction modalities are endless.
Incredible work!
Today, we're sharing new research on GWM Worlds 2. Our latest General World Model that generates interactive real-time simulations in continuous 720p video at 24 fps with 48000 Hz audio.
Built on top of our foundational audio-visual generation model, GWM Worlds 2 allows you to define the environment, subjects, visual style, physical rules and ambience of any world. Once inside, you steer it with text actions addressed to any subject or to the scene itself, alongside continuous camera motion.
The world continues from each new input instead of following a fixed clip or script so sessions have no preset length. GWM Worlds 2 creates a new foundation for interactive entertainment, virtual characters, robotics and embodied-agent simulation.
Learn more about GWM Worlds 2 at the link below.
We are sharing a major update to our General World Models efforts. GWM Worlds 2 can generate full, interactive, real-time video simulations in one continuous 720p stream at 24 fps and audio at 48000 Hz, responding to your inputs as you explore.
It is the first world model that generalizes to arbitrary actions rather than a fixed set of actions or just navigation. Dynamic actions are what actually matter for building gaming and interactive experiences, and for training agents in these simulations.
We also introduced WorldPrompt, a way to tell GWM-2 that a world can be split into two kinds of state: what persists and what changes over time. For example: persistent gravity, physics, and light, alongside actions that describe movement, gestures, object interactions, speech, and sound.
./spiral {Gene Kogan}
2017 to 2022 were quite a special five years for a lot of artists working in this field, with all the new tools, the excitement around creative coding reaching a much wider audience, and of course AI. And it wouldn't be an exaggeration to say that with @ml4a_, @genekogan taught a generation of artists how to work with machine learning.
This year felt like the first time, in ourselves and around us, that enough time had passed to look back and realise it isn't the near past anymore. That's what makes you want a checkpoint, a way to measure the time, the practice and everything that moves around them.
So having Gene in the show made complete sense, and it would have felt like something missing without him. Really glad to have his work in ./spiral. Deep dive in the post below.
One last Easter egg for anyone who wants to make films like these...
All 32 works are forkable and you can have Claude / Codex / Hermes / OpenClaw / your agent of choice making variations of them or totally new pieces with a one-line prompt, no dependencies or extras to install.
Click the "fork" button on any of the videos, and it will bring up a prompt that looks something like: "Generate a variation of this music video using only code. Start from https://t.co/AM4HHTIqmU"
Modify it however you wish! e.g. "add glitch effects", "put a rap track over it", or just tell it to improvise and surprise you.
All the works are listed here: https://t.co/DvqQJGZwMR
New work!
Player Computer is a collection of 32 short musical films exploring a variety of absurdist and surrealist themes.
Inspired by the 19th century player piano, the original art machine, each film is generated frame-by-frame from a single forkable Python script.
Player Computer is being exhibited and sold at the upcoming ./spiral exhibition by @SuperRare this week at Fondazione Aversano in NYC, alongside 10 other amazing artists.
Many thanks to @im_jonooo and @tyaagnliu for producing this!!
Go see it! https://t.co/oU6sON3vxS
First he handed machine learning to artists. Then he began handing agency to the artwork itself.
For ./spiral, @genekogan presents Player Computer, an homage to the player piano built for code.