Sora
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 5
- ByteDance world model demo surfaces, reportedly overseen by Zhang Yiming
A Xiaohongshu post circulating in Chinese tech circles says Tim of the filmmaking channel MediaStorm demonstrated ByteDance's latest world model, which could build a world and let users interact with it — freely controlling camera moves, stepping through a virtual 'anywhere door', and letting robots learn manipulation or doctors train by simulation; ByteDance has not announced the model and the details are unverified. Earlier reports said ByteDance is developing a real-time spatial video generation model based on its Seedance video model, personally coordinated by founder Zhang Yiming across research, cloud infrastructure, Pico hardware and content teams with priority computing support; the model generates 20 fps video on demand with roughly 50ms end-to-end latency, rendering in the cloud and streaming to devices, with planned Pico headset support. World models were reportedly listed first among ByteDance's four 2026 AI priorities, with a goal of shipping at least one model benchmarked against Google's Genie 3 by year-end for livestreaming, interactive short dramas, games and simulation for robots and job training.
钛媒体 · 🔥 4 - Tian Keyu's stealth world-model startup raises ~$30M as poisoning allegations resurface
A Chinese long-form commentary examines the case of Tian Keyu, a former ByteDance intern whom the company fired in August 2024 and later accused of tampering with colleagues' model-training code; ByteDance said in an October 2024 statement that reports of 8,000+ GPUs and tens of millions of dollars in losses were heavily exaggerated. A GitHub page posted the same day as his podcast appearance, signed by five named people including co-first authors of the video project Goku and the first author of LlamaGen, alleged the interference began in March 2024 and included corrupted text-encoder weights, a modified optimizer, altered data shuffling and a shared-storage backdoor; the page also claimed the VAR text-to-image project served PixArt, an open-source model, for daily tests. Tian, who won a NeurIPS best-paper award for VAR, has since raised close to $30 million from 5Y Capital, IDG and others at a roughly $200 million post-money valuation for a still-unnamed world-model company of about ten people; a court reportedly found him in breach of contract and ordered 500,000 yuan in damages.
虎嗅 AI · 🔥 0 - OpenAI in talks to license Sora to Katzenberg-backed filmmaking AI startup
The Information reported on Oct. 12 that OpenAI is negotiating with a stealth AI video startup focused on film production to hand over the shuttered video model Sora, which the startup would adapt and sell as software or services to filmmakers. The startup was founded in recent months by Hollywood producer and DreamWorks co-founder Jeffrey Katzenberg, former Sora research lead Bill Peebles and former Dropbox CFO Sujay Jaswa; its name has not been disclosed. People familiar with the talks said OpenAI would receive an equity stake in the startup if a deal is reached, though the size is unknown and it remains unclear whether the startup would get Sora's model weights. Sora and its TikTok-like app were shut down this year as OpenAI cut server-intensive side projects, and OpenAI also ended a roughly $1 billion Disney partnership; WndrCo, co-founded by Katzenberg and Jaswa, was previously reported to be in talks for a large funding round including a16z.
The Information · 🔥 0 - Netflix's ~$587M buy of Ben Affleck's AI startup InterPositive
InterPositive, the film-AI company Ben Affleck founded in 2022, was acquired by Netflix in March 2026, with the team joining the streamer and Affleck staying on as a senior adviser; an SEC 10-Q filing later showed a roughly $587 million cash acquisition that month. Affleck has since detailed the approach in a Bloomberg Screentime appearance and a GQ interview: the team spent about eight months capturing a proprietary dataset in controlled stages using cameras, LiDAR, tracking gear and virtual-production setups, so models first learn filmmaking grammar — lens language, lighting, editorial continuity — and are then adapted on each project's own dailies, an approach TheWrap describes as a per-film "mini-model" used to fill missing shots, recompose, fix lighting and replace backgrounds. He confirmed his upcoming Netflix film "Animals" used the tools in several stages but declined to say which shots were processed.
MIT科技评论中文 · 🔥 0 - ArtCraft launches an AI image and video creation IDE
ArtCraft, which appeared on GitHub Trending, describes itself as an "intentional crafting engine" for artists, designers and filmmakers, and an IDE for interactive AI image and video creation. It combines 2D compositing, 3D scene staging, character posing and identity transfer, image-to-3D mesh and background removal with text-to-image, prompted image editing and image-to-video flows, letting users pick their own models. Its listed catalog totals 62 models — 16 image, 25 video, 5 music and sound, 11 3D mesh and 5 world/Gaussian-splat models — with integrations including Grok, Midjourney, Sora and World Labs Marble; some entries are marked disabled or restricted, and Kling, Google, Runway and Luma integrations are listed as planned.
GitHub Trending（每日） · 🔥 0
Experience and discussion from the community
Share my Sora experienceAsk about Sora
Nobody has shared their experience with Sora yet.