Good morning. Alibaba dropped HappyHorse 1.1 while Sora is gone and Seedance is frozen. Snap spun off its entire AI video team into a new company called Dotmo. And a real-life "Bane" showed up in a courtroom in handcuffs and nobody could tell if the footage was real or AI. Have you tried HappyHorse yet? Hit reply and tell me. We cover all three below.
🐴 HappyHorse 1.1 Just Dropped and the Timing Is Everything
Alibaba released HappyHorse 1.1 this week with full API access on Alibaba Cloud Model Studio and introductory launch pricing for the first two weeks. HappyHorse 1.0 already topped the Artificial Analysis leaderboards in both text-to-video and image-to-video. Version 1.1 delivers what Alibaba calls production-ready video synthesis across core content creation scenarios.
The timing matters more than the specs. OpenAI shut down its Sora standalone app because the compute costs were unsustainable. ByteDance indefinitely shelved the international rollout of Seedance 2.0 after copyright complaints from Hollywood studios. For enterprise teams that were building workflows around those tools, the competitive landscape just contracted sharply. HappyHorse is stepping into the gap with enterprise SLAs, regional compliance, and aggressive pricing.
For creators, HappyHorse is worth testing now. It's available on Alibaba's own platform, which means you don't need to commit to one ecosystem. The model generates synchronized audio and video in a single pass with native lip sync across seven languages. If you've been locked into Seedance or were waiting on Sora, this is the alternative that's actually shipping.
Snap Just Spun Off Its Entire AI Video Team Into a New Company
Snap spun off its internal generative AI video team into a separate company called Dotmo. The stated reason: the costs of developing AI video models internally were too high. Dotmo will focus on AI models for interactive gaming experiences, not social media. Snap's CTO Bobby Murphy is the lead personal investor and will hold a significant stake in the new company while staying full-time at Snap.
Snap keeps a large equity stake in Dotmo and is licensing its technology to the new company. Dotmo may seek outside funding later. This is Snap's second spinoff this year after separating its Specs smart glasses line into its own entity. Snap also cut about 1,000 jobs earlier in 2026.
For creators, the signal is blunt. Building AI video generation tools from scratch is so expensive that even a company with 950 million daily active users can't afford to do it internally. The companies that survive in AI video will either have massive cloud infrastructure behind them (Google, Alibaba, ByteDance) or raise billions independently (Runway, xAI). Everyone else is spinning off, shutting down, or partnering. The middle ground is disappearing.
A Real-Life "Bane" Showed Up in Court and Nobody Could Tell If It Was AI
A video of a massive man in handcuffs being escorted through a courtroom went viral this week. The guy is built like a comic book villain. Multiple handcuffs. Officers flanking him. The physique is so exaggerated that the first reaction across Instagram and TikTok was "this has to be AI." Compilations hit millions of views. Comment sections turned into debates.
The footage shows a muscular person in a real courtroom. But the fact that the default response to seeing something unbelievable is now "is this AI?" tells you everything about where we are. A year ago, people saw wild footage and assumed it was real until proven fake. Now the assumption is reversed.
This is the same cycle as the bear on the Ring camera, the orangutan saving the toddler, and the goddess fans at baseball games. Every time real footage looks too dramatic, too cinematic, or too perfectly framed, the audience assumes AI first. For creators, this creates a paradox. The better AI video tools get, the less people trust anything they see online. Your real content now has to prove it's real. That's a new problem, and it's not going away.