🎬 AI × Entertainment Daily
Saturday, April 4, 2026
Your daily briefing on the convergence of artificial intelligence and the entertainment industry — covering film, music, gaming, and streaming.
🔥 TODAY’S TOP STORY
1. Netflix Open-Sources VOID — Its First-Ever Public AI Model Rewrites Physics in Video
Netflix has released VOID (Video Object and Interaction Deletion), its first publicly available open-source AI model, published April 3, 2026 on Hugging Face under an Apache 2.0 license. VOID does not merely erase objects from video — it re-simulates the physical consequences of their absence. Remove a person holding a guitar, and the guitar falls. Remove a pedestrian from a crash scene, and the debris, smoke, and flames disappear, replaced by pristine pavement.
VOID is a vision-language model that can not only erase objects from a scene but also inpaint how remaining objects in the scene should behave without the influence of whatever was excised. In human evaluations across multiple scenarios, VOID was preferred 64.8 percent of the time, with Runway coming in a distant second at 18.4 percent.
VOID is built on top of Alibaba’s CogVideoX video diffusion model, fine-tuned with synthetic data from Google’s Kubric and Adobe’s HUMOTO for interaction detection. Google’s Gemini 3 Pro analyzes the scene and identifies affected areas, while Meta’s SAM2 handles segmenting the objects that need to be removed.
The model targets professional film editing and automated video-manipulation workflows, and its open-source release means studios and indie filmmakers alike can deploy it commercially.
| 📎 Source: The Register | The Decoder | Hugging Face |
🎬 FILM & STREAMING
2. OpenAI Kills Sora — Disney Pulls Its $1 Billion Deal
In one of the most dramatic reversals in the AI video race, OpenAI announced on March 24 that it would discontinue Sora, providing no public explanation for the decision. Its consumer app will go dark on April 26, with the API following on September 24, 2026.
The financial picture is damning. At peak usage, Sora burned an estimated $15 million per day in inference costs. Against that, Sora generated just $2.1 million in total lifetime revenue. Disney, which had announced a $1 billion partnership with OpenAI in December 2025, subsequently pulled out — and no money ever changed hands between the two parties before the deal collapsed.
Google moves swiftly to fill the void (no pun intended), having launched Veo 3.1 Lite — priced at under 50% of its Fast tier — specifically targeting the developer market that Sora will soon vacate.
📎 Source: Winbuzzer
3. Google Vids Opens Veo 3.1 to Everyone — Free AI Video Is Now a Google Product
Google Vids now offers any Google account holder up to 10 free AI video clips per month via Veo 3.1. The update, which went live on April 2, 2026, makes Vids one of the first major productivity tools to bundle AI video generation, music creation, and controllable avatars into a single platform with a free starting tier.
For power users, the upgrades go further. Google AI Ultra and Workspace AI Ultra accounts can now generate up to 1,000 Veo videos per month. Google AI Pro and Ultra subscribers can now generate everything from short 30-second clips to three-minute tracks using the integrated Lyria 3 music generation engine.
The platform also introduces directable AI avatars — users can now give characters natural language stage directions, dress them, and place them against custom backdrops. Completed videos can be exported directly to YouTube, defaulting to private for review before publishing.
Google now allows users to export Vids projects directly to YouTube. Videos uploaded this way are set to private by default, preventing premature public releases.
| 📎 Source: Google Blog | Chrome Unboxed |
4. AI Film Restoration Sparks Industry Debate: Wizard of Oz, Magnificent Ambersons in the Crossfire
The ethics of AI touching cinema’s sacred canon have erupted again. An AI-altered version of The Wizard of Oz on display at the orb-shaped Sphere venue in Las Vegas has stoked both curiosity and revulsion — especially from filmworld purists.
Meanwhile, Fable Studios’ Edward Saatchi is using generative AI to reconstruct the 43 minutes of The Magnificent Ambersons famously cut and destroyed by RKO in 1942. One “shoot” with real actors has already been completed, with the missing shots re-created; there will be two additional filming sessions with performers, whose work will be superimposed onto the original actors’ likeness in the film with the help of AI.
The Orson Welles estate, for its part, is firmly opposed. Beatrice Welles stated she is “quite terrified of AI and in many ways wish it had never been invented.” The project nonetheless continues, with Saatchi receiving outreach from established directors he says he cannot yet name.
Experts expect producers to plumb the film archives for other potential hits and enhance them with AI before screening them in venues as varied as IMAX theaters and Cosm, another 360-degree dome with locations in Los Angeles, Dallas and Atlanta.
📎 Source: The Hollywood Reporter
5. Hollywood’s AI Adoption Is Percolating From the Bottom Up — Starting With Assistants
A new Hollywood Reporter investigation reveals that the real story of AI in Hollywood is not Terminator-style replacement, but quiet, mundane integration by the industry’s most junior staff.
The reality of how AI is currently being integrated into the largest swaths of the entertainment industry is much more mundane. And, as with previous introductions of new technology into Hollywood, from digital film to email, AI is percolating from the bottom up, starting with the assistant class — on track to become industry standard as today’s underlings rise to positions of power.
In January, Disney held an internal AI summit with representatives from all departments to promote the integration of AI use across the entire company’s business, from Imagineering to business affairs and beyond.
Hollywood support staff are using AI for composing emails, scheduling, and navigating the relentless stream of congratulatory gifts — but also increasingly for script coverage and creative development, a fact some senior management would rather not acknowledge.
📎 Source: Hollywood Reporter / Yahoo Entertainment
🎵 MUSIC & AUDIO
6. ElevenLabs Quietly Launches ElevenMusic — AI Music Generation Goes Mobile
Voice AI powerhouse ElevenLabs has entered the AI music space. ElevenLabs has quietly released an iOS app called ElevenMusic that can be used to create music with AI and discover AI-generated music in order to compete with platforms like Suno and Udio. The new app, finally released on April 1, suggests ElevenLabs wants to be more than just a voice model company.
ElevenMusic is currently free to use, and users can generate up to seven songs per day using natural language prompts. Users can also adjust the length of the song, if the song has lyrics, and the writing style.
The move signals ElevenLabs’ strategic pivot: the company sees multi-modal AI content creation — not just voice synthesis — as its defense against the eventual commoditization of audio AI.
📎 Source: TechCrunch
🎮 GAMING
7. Take-Two Fires Its Head of AI — Weeks Before GTA 6 Launch
In a move that sent shockwaves through the games industry, the former Head of AI at Take-Two Interactive revealed they had been laid off. The company, which is the parent organization of GTA 6 developer Rockstar, reportedly made a major shake-up in the AI department, laying off an unknown number of employees.
Luke Dicken wrote: “Over the past 6 years — first leading Zynga’s game services division in central tech, then spending the last 2 years building out the AI team at Take-Two — we pushed the edge of what’s possible, applying emerging technologies to real game development problems. That chapter has come to a close, as shifting priorities from upper management have impacted my team and me.”
The irony is hard to miss. CEO Strauss Zelnick had just told investors in the company’s Q3 earnings call that Take-Two is “actively embracing generative AI” and running “hundreds of pilots and implementations across our company.” The cuts come with GTA 6 — arguably the most anticipated game in history — set for a November 19, 2026 release.
| 📎 Source: Insider Gaming | Game Developer |
📺 HARDWARE & HOME ENTERTAINMENT
8. Samsung’s 2026 OLED TVs Put an AI Companion in Every Living Room
Samsung’s 2026 television lineup represents a fundamental shift in how displays integrate with our digital lives, introducing Vision AI Companion for contextual intelligence, 165Hz gaming capabilities, and refined Frame-style designs that blend displays into living spaces.
The Vision AI system is the standout: it integrates with Samsung’s SmartThings ecosystem to control smart home devices, learns household routines, and can automatically dim lights when a movie starts. On the gaming side, gamers get a serious upgrade with Motion Xcelerator 165Hz, along with G-SYNC and FreeSync Premium Pro support — Samsung’s play for the premium gaming-beyond-consoles market.
The 2026 lineup reframes the television not as a passive display but as the AI hub for an entire living space.
📎 Source: Sammy Fans
🌏 INDUSTRY SIGNALS
9. Asia Moves Fast: Filmart Showcases Region’s Full-Throated AI Embrace
While Hollywood remains tied in knots over labor contracts and copyright litigation, the mood could not be more different in Asia. As Hollywood remains locked in labor and legal battles over generative AI, Filmart is showcasing Asia’s increasingly full-throated embrace of the technology as both a foregone conclusion and the industry’s next growth engine — with 28 talks devoted to the subject this year.
The contrast is stark: Asian studios and streaming platforms appear to be treating generative AI as infrastructure, not controversy.
📎 Source: The Hollywood Reporter
10. Google Veo 3.1 Lite Gives Developers a Cost-Effective Entry Point for AI Video Apps
Beyond the consumer-facing Vids update, Google is making a play for the developer ecosystem. Google is introducing Veo 3.1 Lite, its most cost-effective video model. This model empowers developers to build high-volume video applications, at less than 50% of the cost of Veo 3.1 Fast, but with the same speed.
The model supports Text-to-Video and Image-to-Video in 720p and 1080p, in both landscape (16:9) and portrait (9:16) orientations. A further pricing cut on Veo 3.1 Fast is scheduled for April 7 — compressing the cost stack just as OpenAI’s Sora vacates the developer API market.
📎 Source: Google Developers Blog
📊 THE WEEK IN CONTEXT
| Theme | Signal |
|---|---|
| AI Video | Netflix VOID + Google Veo 3.1 free tier = video AI democratization accelerates |
| Sora Shutdown | OpenAI’s costly retreat opens space for Google, Runway, and open-source models |
| Gaming | Take-Two AI layoffs signal a recalibration, not a retreat, in games industry AI spend |
| Music | ElevenLabs’ mobile move means AI music generation is now a phone-first product |
| Film Preservation | AI restoration debates heating up — cultural and legal questions unresolved |
💡 EDITOR’S TAKE
The most significant story this week is not the one generating the most clicks. Netflix releasing VOID — its first-ever open-source model — is a quiet earthquake. It signals that streaming giants are no longer just consumers of AI tools; they are now contributors to the open ecosystem that will define how film and video is made for the next decade.
Meanwhile, the OpenAI Sora shutdown is a cautionary tale that no amount of hype can escape: inference economics matter. At $15M/day burn against $2.1M lifetime revenue, Sora’s collapse is a reminder that the AI entertainment gold rush still has a very real gravity.
