Image created with gemini-2.5-flash-image with claude-sonnet-4-5. Image prompt: A miniature film studio set capturing a chess game with stop-motion animation equipment, a white knight piece frozen mid-leap with motion blur trails, tiny spotlights and camera rigs surrounding the board, film reels and clapperboards labeled with AI timestamps in the foreground, cinematic lighting with dramatic shadows, shallow depth of field.
I am continually surprised about how few applications take advantage of the fact that AI systems can work with video. For example, I can ask Gemini questions about what happens in a video (and not mentioned in a transcript) and get coherent answers including identifying emotion https://x.com/emollick/status/1980695990790418889
SAG-AFTRA, OpenAI, Bryan Cranston Collaborate to Ensure Voice and Likeness Protections in Sora 2 | SAG-AFTRA https://www.sagaftra.org/sag-aftra-openai-bryan-cranston-collaborate-ensure-voice-and-likeness-protections-sora-2
Netflix is betting that AI augmentation beats AI replacement, and they might be fighting the last war. There’s basically two ways generative AI eats Hollywood: 1. The Netflix path – slip AI tools into the existing pipeline. Make VFX 10x faster, de-age actors without breaking https://x.com/bilawalsidhu/status/1980903106943885694
It makes perfect sense to let agents understand, imitate, and learn how humans use computers from videos! We present VideoAgentTrek, which builds strong computer-use agents through video pretraining and agentic tuning. This approach has already proven effective in the training of”” / X https://x.com/huybery/status/1981728838024560669
AI video models may not be complete world models, but they are oddly capable of fairly sophisticated (if flawed) “”simulations”” of novel situations. Veo 3.1: “”three toy ships, one made of iron, the other of wood, and one out of loosely packed sugar, fall into a pool of water”” https://x.com/emollick/status/1980126684306424155
Ethan Mollick on X: “This is not perfect, but it also doesn’t seem like a model trained on video should be able to get so many details of the dynamics right: “honey pours down a marble statue of a toucan, the nose cracks and falls off” https://t.co/ei0I5ecmom” / X
https://x.com/emollick/status/1980128284294938661
Veo is getting new precision editing capabilities that let you easily add or remove elements from a scene – all while preserving the integrity of your original video. 🎥 https://x.com/GoogleDeepMind/status/1980261047836508213
Sora 2; “a movie trailer whose genre you can never quite figure out because it keeps drawing on cliches from many different genres through a series of fast cuts, make it over the top, featuring movie guy voice” https://x.com/emollick/status/1980285871002657026
Datasets you need to build an AI JARVIS — Meta dropped 500 hours of 3D motion data spanning everything from individual gestures to multi-person conversations and co-living scenarios, complete with motion tracking, annotations, and audio tracks. https://x.com/bilawalsidhu/status/1980719297669525925
Now this is fking cool. Genie 3 made the World Labs approach look dated. But now they drop a step towards the best of both worlds. Generative video w/ spatial memory. And a live demo to boot. Making the bitter lesson sweet. Nicely done World Labs team! https://x.com/bilawalsidhu/status/1978917822228045858
Virtually Being https://eyeline-labs.github.io/Virtually-Being/
RTFM, a new generative interactive world model by World Labs, generates real-time video from a single image for exploring 3D worlds. Trained on large-scale video data to predict the next frame. Achieves unbounded persistence via spatial memory. try it: https://x.com/TheHumanoidHub/status/1978878123275096579
腾讯混元3D https://3d-models.hunyuan.tencent.com/world/
Today, we are open-sourcing Hunyuan World 1.1 (WorldMirror), a universal feed-forward 3D reconstruction model. 🚀🚀🚀 While our previously released Hunyuan World 1.0 (open-sourced, lite version deployable on consumer GPUs) focused on generating 3D worlds from text or https://x.com/TencentHunyuan/status/1980930623536837013
So these researchers figured out you can basically hallucinate 3D cities into existence using just satellite photos & a diffusion model. The problem’s pretty straightforward: satellites only see rooftops. Building facades? Invisible. Street-level detail? Doesn’t exist. But https://x.com/bilawalsidhu/status/1981036580153544854
New Gemini Drops just dropped! Create richer videos with Veo 3.1, generate slides in Canvas, get personalized recommendations on Google TV, and more. Catch up on all the latest Gemini news from this month 🧵 1/9 https://x.com/GeminiApp/status/1981760415580528901
Google presents VISTA A Test-Time Self-Improving Video Generation Agent https://x.com/_akhaliq/status/1980398215707906391
🚨🎬 Big news from Video Arena! @GoogleDeepMind’s latest Veo 3.1 now ranks #1 in both Text-to-Video and Image-to-Video leaderboards. 🏆 This is a +30-point leap from Veo 3.0 → 3.1, making it the first model to break 1400 in Video Arena history! Huge congrats to the https://x.com/arena/status/1980319296120320243
I like light probes and I cannot lie. Especially when they’re being hallucinated by AI. Simulon is building a much higher quality version of image-based lighting estimation the Vision Pro or ARCore does in real time.”” / X https://x.com/bilawalsidhu/status/1979291609230659947
Adobe launches a foundry service that builds custom generative AI models for enterprises | TechCrunch https://techcrunch.com/2025/10/20/adobe-launches-a-foundry-service-that-builds-custom-generative-ai-models-for-enterprises/
Meta AI’s app downloads and daily users spiked after launch of ‘Vibes’ AI video feed | TechCrunch https://techcrunch.com/2025/10/20/meta-ais-app-downloads-and-daily-users-spiked-after-launch-of-vibes-ai-video-feed/
The new Sora Pro feature that builds storyboards and executes them is really interesting. Here is the prompt “an ad for the abstract concept of the feeling you get after falling asleep on your arm” Notice the high character consistency, multiple shots, composition. All the AI. https://x.com/emollick/status/1979020908892176718
Statement from OpenAI and King Estate, Inc. The Estate of Martin Luther King, Jr., Inc. (King, Inc.) and OpenAI have worked together to address how Dr. Martin Luther King Jr.’s likeness is represented in Sora generations. Some users generated disrespectful depictions of Dr.”” / X https://x.com/OpenAINewsroom/status/1979005850166648933
Announcing Lindy AI CMO, a team of agents running entire marketing workflows end-to-end, handling market research, analysis, and creative at scale — so you can start thousands of ad experiments in minutes. As part of this, we’re launching new integrations with Sora 2, Veo 3.1, https://x.com/Altimor/status/1978869339974701299
Sora 2 Pro is very impressive in my tests but really needs its own interface – using a storyboard maker shoved inside a video creation tool shoved inside of a drafts folder of a social media app is awkward.”” / X https://x.com/emollick/status/1980132516406473126
Introducing: Interactive Sora! A choose-your-own adventure GAME powered by Sora 2. Every choice spins up a brand-new scene instantly. Open source, link here: https://x.com/mattshumer_/status/1978848940083839162
Introducing LTX-2: the most complete open-source AI creative engine. – Synchronized audio and video generation – Native 4K fidelity, up to 50 fps and 10 s+ sequences – API-first design for seamless integration into creative pipelines – Runs efficiently on consumer GPUs – Fully https://x.com/ltx_model/status/1981346235194683497
World-in-World: World Models in a Closed-Loop World https://world-in-world.github.io/
MoGA Mixture-of-Groups Attention for End-to-End Long Video Generation https://x.com/_akhaliq/status/1980952993563349127
🎬 Video generation is now available on Together AI Through our partnership with @Runware, we’re integrating 20+ video models (Sora 2, Veo 3, PixVerse V5, Seedance) and 15+ image models All through the same APIs you use for text inference → https://x.com/togethercompute/status/1980746093932515697
TV in the 90s: you turn it on, you watch. TV 2025: – turn on, wait for it to load – popup: TV wants to update, 1.5GB. No. – scroll sideways, find prime video app or etc – popup: now app wants to update, 500MB. No!! – App launching… App loading… – select account screen – 🫠”” / X https://x.com/karpathy/status/1978653908663726585
Introducing Workflows, a new way to build your own tools inside of Runway. Now you can create your own custom node-based workflows chaining together multiple models, modalities and intermediary steps for even more control of your generations. Build the Workflows that work for https://x.com/runwayml/status/1980736639405289786
Introducing “Argil Atom” the most controllable video AI model in the world! We cracked two core challenges, setting a new SOTA in: Consistency and Control with no limit on video duration. RT + reply 🍿 within 10h = 300 credits in your DMs 👀 https://x.com/BrivaelLp/status/1981343140196778270
HoloCine Holistic Generation of Cinematic Multi-Shot Long Video Narratives https://x.com/_akhaliq/status/1981561283737456898
Today we’re launching LTX-2 in LTX (we’re more than a studio). LTX-2 represents a major breakthrough in speed and quality. It offers fast generation, native 4K, lipsync, high frame rate, and up to 10s of continuous video. Here’s what you need to know 🧵 https://x.com/LTXStudio/status/1981371951894667279
Today we’re introducing the Apps for Advertising collection. A growing library of easy-to-use apps inside of Runway that allow you to get from idea to deliverable with less time, money and effort. From generating endless product shots for social and e-commerce to mocking up https://x.com/runwayml/status/1981380360249159783
You can just transfer your entire performance to any character now. Fed it my guitar playing to drive this avatar, and it holds up well. Wan 2.2 Fun Control and Animate are powerful primitives. At the rate things are going, this’ll be a real-time AI filter. And at that point https://x.com/bilawalsidhu/status/1978871705599905892
Veo 3.1 Start Frame -> End Frame is actually overpowered 🤯 Look at the extreme detail we were able to bring into this cellphone footage and then seamlessly blend back! We also used our Veo 3.1 Prompter to write this prompt 😉 Try them both below 👇 https://x.com/heyglif/status/1980362634982748332
Ruining more great art with Veo 3.1 (sound on) https://x.com/emollick/status/1980863465226113115
Awesome to see Veo 3.1 top the LMArena video leaderboards by a large distance with big improvements over Veo 3.0 for text-to-video (+30) and image-to-video (+70)! 🔥Huge congrats to the team! Try it for yourself in https://x.com/demishassabis/status/1980397419658645708
today we’re open-sourcing Krea Realtime. this 14B autoregressive model is 10x larger than any open-source equivalent, and it can generate long-form videos at 11 fps on a single B200. weights and technical report below 👇 https://x.com/krea_ai/status/1980358158376988747
krea/krea-realtime-video · Hugging Face https://huggingface.co/krea/krea-realtime-video
product storyboards at this quality didn’t exist before 🍿 prompt/reference -> 8 cinematic frames with locked consistency & full directorial control Higgsfield Popcorn made professional AI storyboarding possible 🍿🔪🍌 the new standard. rt + reply in 9h = 225 credits in DM https://x.com/higgsfield_ai/status/1981865084231331921
Qwen just released Qwen3-VL on Hugging Face The most powerful vision-language model in the Qwen series, with comprehensive upgrades across text understanding, visual reasoning, and long context video analysis. From GUI operations to 1M context. https://x.com/HuggingPapers/status/1980809413045940553
Open-o3 Video Grounded Video Reasoning with Explicit Spatio-Temporal Evidence https://x.com/_akhaliq/status/1981564465897509333
Ditto Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset https://x.com/_akhaliq/status/1980265202500116525
UltraGen High-Resolution Video Generation with Hierarchical Attention https://x.com/_akhaliq/status/1980952631544799705
