Video Generation

Today's Stories

1. 1

   Apple researchers tackle text-to-video-with-audio generation challenges

   Apple researchers published a study addressing two key problems in text-to-sounding-video generation—a process that creates synchronized video and audio from text. The research identifies text conditioning as a bottleneck and proposes solutions for cross-modal feature interaction. Generating videos with aligned audio from text descriptions remains difficult because shared captions can cause interference between audio and video, and there is a gap between training captions and the shorter prompts users actually type. Better solutions here could improve how AI tools handle joint audio-video creation.

   The study proposes specific approaches to address text conditioning and modal fusion, though the article does not specify when or how these techniques will be deployed in products or released to developers.

2. 2

   Soracom Transforms into an After AI Organization in One Year, Becoming a Safe Container for "Token Capital"

   Soracom Transforms into an After AI Organization in One Year, Becoming a Safe Container for "Token Capital"

3. 3

   AI ROI Outside Tech May Take Years to Materialize

   A recent analysis suggests that return on investment for AI adoption may take significantly longer to materialize in non-technology sectors compared to the tech industry itself. Businesses investing heavily in AI infrastructure outside of core technology operations face uncertainty about when they will see measurable financial returns, which could affect capital allocation decisions and investor patience with AI spending.

   The timeline and visibility into when non-tech sectors will achieve measurable AI-driven productivity gains or cost savings, as this will determine whether enterprise AI investments remain a priority.

4. 4

   OpenAI shuts Sora video app, Disney deal scrapped

   OpenAI announced Tuesday it is shutting down the Sora app, the video generation tool the company launched in December 2024. The $1 billion（約1600億円） partnership deal with Disney, announced in December 2025 to let users create videos with Disney characters, is also being terminated. OpenAI did not confirm reports that Sora would move into ChatGPT, but promised to share timelines for the app and API soon. Sora faced steep competition from other AI video tools like Google's Veo and Luma Ray and never captured the early momentum from its launch. The sudden exit and Disney deal cancellation, just months after CEO Bob Iger was promoting the partnership in February 2026, signal internal turmoil at OpenAI—a company also dealing with wrongful death lawsuits tied to teen mental health concerns.

   OpenAI has promised to share more details soon, including timelines for preserving user work. The deal was originally set to last three years; Disney said it "respects" OpenAI's decision to exit video generation and shift priorities elsewhere.

5. 5

   Apple releases VideoFlexTok, a new video tokenization method

   Apple researchers introduced VideoFlexTok, a technique that compresses video data into tokens—simplified digital units—in a way that adapts to how complex different parts of a video are. Unlike the standard approach, which treats all video equally, VideoFlexTok uses fewer tokens for simpler scenes and more tokens for complex ones. Video tokenizers determine what details a downstream AI model (like a text-to-video generator) must learn and predict. The conventional method requires models to predict all pixel-level details regardless of scene complexity, which is computationally wasteful. VideoFlexTok may allow AI models to focus computational effort more efficiently, potentially reducing the work required to generate or process video.

   The research was published on Apple's machine learning blog, signaling the company's contribution to AI video understanding. The method's practical adoption by other researchers and whether it influences how future video AI models are built will indicate its real-world impact.

6. 6

   AI Music Video Generation: 10 Tools That Automate Your Creative Workflow in 2026

   AI Music Video Generation: 10 Tools That Automate Your Creative Workflow in 2026