Hosting a worshop on the new Overworld video world model. Come join us! x.com/yondonfu/statu…

x.com/i/article/2014…

Monday's watercooler gonna hit different 💧 The XMTP team is coming through to talk about what we're building together: Real-time video + interoperable messaging = the foundation for interactive media on the open internet. Jan 26, 3 PM ET on Discord.

Livepeer on Arbitrum: Seamless Video for Modern Apps Streaming video is everywhere, but integrating live or on-demand video into your app can feel complex. Handling encoding, playback, and APIs often slows developers down. Livepeer on Arbitrum solves this by combining

Last call for the Daydream Interactive AI Video Program. 2 weeks. 15 creators. Building the future of interactive video before it's mainstream. If you're already experimenting with real-time AI video or want to be, this is where you start. Apply now ↴

Each new cohort in the Interactive AI Video program has been pushing the boundaries on what's possible with realtime AI. Check out this video for a preview, and join the next cohort!

Kova 🤝 Livepeer We’re bringing Livepeer’s decentralized video transcoding and AI inference utility onto Kova. Unlocking real, production-ready workloads on-chain.

Watercooler tomorrow @ 3pm ET 💧 Good time to talk Livepeer updates, what you're building, or whatever's on your mind about real-time AI video. Drop in if you're around ↓

Dive into the Q4 report on the state of Livepeer

📁 real-time AI video └📁 things you can build now └📁 experiences that weren't possible before └📁 folder is too large to open

One year of Livepeer by the numbers 💚

The best real-time AI video applications haven't been built yet. @DaydreamAI's hackathon is how we discover them. Applications closes tomorrow - apply now if you haven’t already ⚡

Self-driving models need rare, dangerous scenarios: pedestrian steps into traffic, heavy rain, dusk. You can spend months capturing this in the real world, or generate unlimited variations synthetically at 50-80% lower cost than centralized clouds. One scales. One doesn't 👀

A quick tutorial for making real time AI visuals using Resolume + Daydream Scope.

Building the backbone for real-time AI video ⚡️

Seedance, Veo 3.1, Kling 3.0, Sora 2 The model race is proving the real-time AI video demand is exploding. But… Who runs the compute when millions of creators want to generate this at scale without paying ByteDance or AWS prices? 👀

Real-time AI video requires: → Speed that matches live camera feeds → Zero noticeable lag → Costs that don't kill your business model Centralized clouds weren't designed for this workload. Livepeer was ⚡️

The $100k/year per agent makes sense as centralized clouds weren't built for continuous AI workloads. They were built to maximize margin. Distributed GPU networks flip this. Orchestrators compete for jobs, no one owns the pricing floor. Costs go down as you scale, not up.

GM What's the most underrated use case for real-time AI video? I'll go first: Fitness apps correcting your form mid-rep. Live shopping where AI demos as you watch. Musicians with visuals that react to their sound. Surgeons getting AI overlays during procedures. Your turn 👇

↰ | $225 billion on inference costs through 2030. | | AI video will scale to hundreds of millions of users | sooner than you think | └ What infrastructure can handle that without costs spiraling out of control?