Video Generation

Today's Stories

1. 1

   AI video displaces actors in China

   Since ByteDance's Seedance 2.0 launched, AI-generated videos are replacing human performers in Chinese productions. In Q1 2026, about 128,000 short dramas were published, three times the total for all of 2025, with ninety-five percent AI-generated, per the China Netcasting Services Association. The industry employs 690,000 people directly, and 15 million list livestreaming as their primary job. Some performers are forced to "distil" their voice and likeness into AI tools before getting fired, and AI-related labor disputes have risen over the past two to three years.

   One minute of AI video now costs just $90 to $120, about ten percent of what human production used to cost, according to Tsinghua University professor Shen Yang. With Seedance 2.5 and Wan 3.0, AI video quality has jumped again.

2. 2

   What Counts as a World Model?

   A Reddit user asked the ML community to clarify what a world model is, noting that most current builds are fancy video generation models. The term is contested, with simulators, physics engines, emulators, and digital twins all being considered as possible world models, which creates confusion for non-experts.

   The discussion hinges on whether a world model must operate on learned representations, not just hand-crafted physics, leaving room for ML-based simulators.

3. 3

   BearJam: AI automation needs human oversight

   BearJam, an AI-powered video production company, shared its approach to responsible automation after two years of experience, emphasizing a hybrid model that combines AI with human judgment. The company argues that automation scales errors as fast as output, so human checkpoints and clear limits are essential to avoid legal and reputational risks, especially since liability for AI output rests with the deployer, not the platform.

   BearJam has turned down projects it could technically deliver, including misleading before-and-after imagery, and commits 1% of revenue from AI-intensive projects to nature restoration in Kent.

4. 4

   Google launches Gemini Omni 1.1 Flash video AI

   Google announced Gemini Omni 1.1 Flash on August 27, a multimodal AI model for video generation and editing, available to developers via the Gemini API and Google AI Studio. The update makes Omni 1.1 production-ready for professional use. The key enhancement is scene extension, which now analyzes up to 10 seconds of prior context (vs. the previous 1 second), improving visual and narrative coherence. Videos can be extended in 10-second increments up to 40 seconds total.

   A 360p draft mode generates up to 60% faster and at one-third the cost of the standard 720p, with final output upscalable to 1080p or 4K. Adobe has already integrated Gemini Omni Flash into Adobe Firefly.

5. 5

   Gemini Omni Flash 1.1 cuts video AI cost, adds long scenes

   Google updated Gemini Omni Flash to version 1.1, improving scene extension to analyze up to ten seconds of video (previously just the last second). Scenes can now be extended in 10-second increments up to 40 seconds, and developers can upload external footage as a style reference. The update makes AI video generation cheaper and more flexible. A new 360p draft mode runs up to 60 percent faster at a third of the cost of 720p, with per-second pricing from $0.03 (360p) to $0.30 (4K). This could lower the barrier for businesses to produce video content.

   Omni 1.1 is available through Google AI Studio and developer docs. Pricing per second: $0.03 for 360p, $0.10 for 720p, $0.15 for 1080p, and $0.30 for 4K. Videos can be upscaled to 1080p or 4K.

6. 6

   Flow AI launches Veo-powered video generator

   Flow AI, an independent platform, launched an online AI video generator powered by Google's Veo technology. It turns text and images into cinematic clips in seconds, with no install or GPU required. This gives non-technical users a quick, accessible way to create video content without specialized hardware or software, potentially lowering the barrier for video production. The platform is independent and not affiliated with Google, using third-party models under their respective licenses.

   The service is online-based, meaning it can be used from any device with internet access. Its reliance on Google's Veo technology may shape its capabilities, though specifics on pricing or availability are not mentioned.