ai video is getting good in a way that feels less like ‘a new tool’ and more like ‘the new default’. you don’t just make a clip, you can iterate like software: change the shot, fix the pacing, re-light it, translate it, spit out 20 versions for 20 audiences, all without a traditional production pipeline. 2026 we will see high quality 10 minute+ short films and shows, no doubt about that. Ai slop will still exist but so will high quality content too. what’s bullish is the signal from the top. disney just did a deal with openai that lets people generate videos using disney / marvel / pixar / star wars characters inside sora. that’s basically “we’re leaning in, not fighting it”. and netflix is trying to pull the other side of the value chain into its orbit. they’ve agreed a $72bn deal to buy warner bros discovery’s studios + streaming assets (pending approvals), which is the clearest ‘we want the biggest libraries + production engine under one roof move yet if ai output keeps stepping up, i think you end up with fewer traditional studios over time. not because stories go away, but because a lot of “studio work” gets absorbed into software + platforms. more content gets made, but the old overhead shrinks this is where the market still feels inefficient: everyone chases the shiny apps and viral demos, but the stuff that compounds is underneath. video is expensive to generate at scale, so value piles up in compute, storage, delivery, and the data layer (clean, rights-cleared, well-organised) that keeps the whole machine running without constant friction ai video will look like magic on the surface. the real winners are the ones owning the pipes and the shovels that make that magic reliable and cheap for everyone else our job at Script Network is to build out and license as much of that pipeline as possible. the data layer for ai content will be the most important part of the ai industry, and 2026 this will really come to light.