Now That OpenAI Has Terminated Sora, How Will I Have This Blog Read To Me By An Uncanny Buxom Mike Wazowski?

I have bad news for anyone who was really into consuming or making videos of Charlie Kirk arresting Jeffrey Epstein, or SpongeBob SquarePants fighting Barack Obama in the format of a side-scrolling fighting game: OpenAI has suddenly and unceremoniously pulled the plug on Sora, its text-to-video generator app, after just four months and before a billion-dollar partnership with Disney ever really got underway. The long-prophesied supplication of all entertainment to uncanny, glitchy videos of Anthony Bourdain smoking weed with Mao Zedong will have to wait.

OpenAI’s first forays into video generation began in late 2024, though it was not until the release of Sora 2 in September 2025 that they wanted the world to take them seriously. More than a million people downloaded the app on Sept. 30, the day of its release; three months later, Disney announced that it had reached a landmark agreement to license its famously copyright-protected cast of animated characters for use in the app. This was a big enough deal that Disney CEO Bob Iger and OpenAI CEO Sam Altman made the rounds together, to boast of the transformative potential of allowing people to generate videos of Nemo the clownfish performing oral sex on Goofy.

As part of the agreement, Disney was set to invest $1 billion in OpenAI. In addition, Disney announced a suite of integrations it planned to shoehorn into its Disney+ streaming app. “Under the license, fans will be able to watch curated selections of Sora-generated videos on Disney+,” the press release boasted, “and OpenAI and Disney will collaborate to utilize OpenAI’s models to power new experiences for Disney + subscribers, furthering innovative and creative ways to connect with Disney’s stories and characters.”

The intent was clear: to obviate existing forms of entertainment which, unlike typing “video of Steamboat Willie flying airplane into Pentagon” into Sora, require employing people to perform labor. Sora’s video generation, by contrast, only requires stealing already performed labor. As no shortage of AI fanboys and exponents immediately claimed, this deal was a supposed death knell for Hollywood: Even one of the richest and most powerful companies in the entertainment industry was bending the knee, acknowledging reality, and hurtling into the user-generated future as quickly as possible. Certainly that’s a future that those in control of the industry would like to live in, as it represents a final triumph of management over labor, and of automation over the messy shared responsibility of relying on talented and creative human beings to write, act, and direct.

Billions of dollars in venture capital have funded several AI products that explicitly aim to automate the entertainment industry. Even if those products have yet to do so, and even if demand among audiences for that type of stuff is at best uncertain, it has certainly spooked Hollywood. The economy around movie and TV production appears to be deflating at an alarming rate, and stories abound of directors, production assistants, and actors having an enormously difficult time finding routine work. Even reality TV, itself a cheap tool engineered to break the labor power of filmmakers and performers, is subject to harsh cuts across the industry.

It is important to consider the Disney deal in that context. Disney and OpenAI should be natural enemies: the vanguard of legitimate film production against the most powerful and avaricious of thieves. The agreement was a hedged concession from Disney, an acknowledgement that even if the stuff Sora was capable of generating was inane, stupid, and in competition with the slick, big-budget stuff Disney produces, the amount of money behind it made it impossible to ignore. Better to reach an agreement instead of waiting to be flattened.

The good news (for the human race, if not for the subset of it who yearn for full-motion renderings of Jack Skellington breastfeeding Buzz Lightyear) is that Sora is such a money-losing piece of shit that OpenAI had to tear up the deal before it ever got going. The video platform is estimated to have lost $15 million per day, which means OpenAI would have burnt through the Disney investment in two months. As recent Wall Street Journal reporting has detailed, OpenAI is now attempting to streamline its cash-furnace business model ahead of a probable public offering in late 2026.

OpenAI lost $12 billion two quarters ago, and internal projections say it will need to torch more than 10 times that amount to scale to profitability. Losing so much money is especially inconvenient right now, as demand from the Persian Gulf states—presently somewhat distracted by Donald Trump’s war of aggression against Iran—has been propping up the industry. OpenAI has lost ground to Anthropic and Google in the AI wars, and the series of consumer-facing products it rolled out over the last year, such as Sora, a web browser called Atlas, and whatever hardware former Apple design guru Jony Ive is cooking up, have mostly scattered the company’s computing power away from its core suite of products and into things people simply do not want. Eliminating Sora is part of a pivot to focusing on business-to-business products, which is where the real potential for growth in AI adoption can be found.

This should be comforting. The death of Sora does not mean that the AI bubble is about to pop right now, nor that the future will be significantly less bleak; every technology company and venture capital firm is still racing to secure control over technology that they hope will hasten wealth consolidation and obviate the need for human labor. But it does mean that none of this is as inevitable as everyone with a stake in convincing you of its inevitability has warned or boasted.

As a recent 404 Media story detailed, nobody likes or is watching the AI sludge intended to compete with real movies and TV. Now the costliest and most hyped effort at expanding the creation of that sludge has been mothballed, due to nobody wanting it. Just because billions of dollars have been invested in destroying human entertainment production, and replacing it with 30-second videos of Yoda shoplifting from WalMart, does not mean the investment will pay off.