A US stealth bomber flies across the darkened sky towards Iran. Meanwhile, in Tehran, a lonely woman feeds a stray cat in a tiled blew of a recent Israeli airstrike.
For beginner viewers, this could be a cinematic retelling of the geopolitical crisis that unfolded just a few weeks ago.
But despite its sophisticated production appearance, it is not shot anywhere and there is no place. And the woman who feeds stray cats is not an actor. She does not exist.
The obsessive footage is “Rough Cut,” a 12-minute short film about an attack on Iranian nuclear sites in the US last month, made by directors Samir Malal and Bukha Kazumi. It is also completely made by artificial intelligence.
The clip is based on details the filmmakers read in news reports of the US bombing. A woman walking along the empty streets of Tehran is a woman who fed a stray cat. Armed with information, they were able to create sequences that looked like they had created by Hollywood directors.
The impressive speed, and for some, the peace of mind of worry that can make this kind of film, has not been lost in broadcasting experts.
Last week, TV producer and bestselling author Richard Osman said that the era of history in the entertainment industry has ended and something new has begun.
“So I saw this and thought, 'Well, that's the end of the beginning of something else and something else,” he said for the rest of the entertainment podcast.
Osman added:
In the case of Mallal, the London-based documentary maker who created ads for Samsung and Coca-Cola, AI provided him with a new form of “Cinematic News.”
The Tehran movie, called the Midnight Drop, is a follow-up to Sky in the Sky, a recreational recreation of Ukrainian drone attacks against Russian bombers in June.
Within two weeks, Maral, who directed the Spiders himself in the Sky, was able to make a film about the Ukrainian attack, which would have cost millions of dollars and would have taken at least two years, including development.
“It should be possible to use AI to create something you've never seen before,” he said. “I have never seen a film news piece that turns around in two weeks. I have never seen a thriller based on news made in two weeks.”
Spiders in the Sky was primarily made with VEO3 with an AI video generation model developed by Google, as well as other AI tools. ChatGpt helped Mallal edit the lengthy interview with the drone operator who formed the spine of the film's story, but Voiceover, Script, and Music were not created by AI.
Google's film production tool Flow is powered by VEO3. It also creates audio, sound effects and background noise. Since its release in May, the tool's impact on YouTube (owned by Google) and social media in general have been marked. As Ottoman's podcast partner Marina Hyde said last week, “The spread is extraordinary.”
Olympic diving dogs have attractive qualities, but quite a lot of them are “slops.”
Maral and Kazumi are aiming to complete a film that stealth bomber missions and thwarts the Iranian story, six-half times the length of a spider in August. It is created by combining models such as VEO3, Openai's Sora, and Midjourney.
“I'm trying to prove the point,” says Malal. “It means you can make something really good on a high level, but it's fast at the speed of culture. Hollywood in particular moves incredibly slow.”
After the newsletter promotion
He adds: “The creative process is making bad things to get good. We have the best bad ideas faster. But the process is accelerated with AI.”
Mallal and Kazmi recently suspended and suspended a short film about another recent news event that appeared on the BBC, The 3i/Atlas Comet.
David Jones, CEO of BrandTech Group, is an advertising startup that uses Generic AI, a term for tools such as chatbots and video generators to create marketing campaigns.
“Today, less than 1% of all branded content is created using Gen AI. 100% of it is created fully or partially using GENAI,” he says.
Netflix revealed last week that it first used AI on one of its television shows.
However, behind this latest surge in AI-spurred creativity is the issue of copyright. In the UK, the creative industry is furious at the government's proposal to train a model of copyrighted work without seeking owner permission, unless the owner opts out of the process.
Maral says he wants to see “a wide accessible and easy-to-use program where artists are compensated for their work.”
Beevan Kidron, a crossbench peer and a leading campaigner against the government's proposal, says AI's filmmaking tools are “great,” but at what point do you realize that these tools are literally based on the work of creators?” she adds: “Creators need fairness in the new system or lose something valuable.”
YouTube has made its terms and conditions that allow Google to use the creator's work to create AI models, and denies that all YouTube stock is used to train the models.
Mallal calls for the use of AI to make it a “promptocraft.” This is a phrase that uses terms to give instructions to AI systems. When making Ukrainian films, he says he was surprised at how quickly the camera angle and lighting tone can be adjusted with a few taps on the keyboard.
“I'm deep in AI. I learned how to encourage engineers. I learned how to translate my skills as a director into prompts. But I've never produced anything creative after that. Then Veo3 came out.
