OpenAI Courts Hollywood in Meetings With Film Studios, Directors
Bloomberg report: https://www.bloomberg.com/news/articles/2024-03-22/openai-courts-hollywood-in-meetings-with-film-studios-directors
Bloomberg alleges that OpenAI is pitching its AI Video Generation tool, Sora, to Film Studios & Directors in Hollywood
OpenAI CEO Sam Altman attended LA parties during Oscars weekend
Unspecified "big name" directors and actors already have access to Sora, the report says.
Chief Executive Sam Altman and Chief Operating Officer Brad Lightcap gave presentations to executives from the film industry giants, said multiple people with knowledge of the meetings, which took place in recent days.
Media analyst Enders said the reception from the movie industry had been broadly optimistic on Sora as it is “seen completely as a cost-saving element, rather than impacting the creative ethos of storytelling.”
The specific details are unknown. A company spokesperson gave Bloomberg a vaguely worded response on its plans: “OpenAI has a deliberate strategy of working in collaboration with industry through a process of iterative deployment – rolling out AI advances in phases – in order to ensure safe implementation and to give people an idea of what’s on the horizon. We look forward to an ongoing dialogue with artists and creatives.”
RESOLUTION
OpenAI will need to announce studios & filmmakers it's partnering with for Sora
The partnerships must be made prior to the release of Sora to the general public
Please comment any suggestions for directors & film studios you'd like me to add
1,000
1.00Jim’s the director who has most embraced breaking technologies across the last 30-40 years. That being said, most of Hollywood’s creative types have gone on the record as really hating AI, Jim included:
@mattyb Wow, yeah I would've expected him to be more in favor of it as well.
It's possible that the people that embrace it will be those outside of Hollywood, much as the early influencers of social media & YouTube were not professional entertainers.
I can understand their concern given that there are only a handful of filmmakers that are able to make high quality films as it's so capital-intensive. Once such capabilities are in the hands of basically everyone it'll also mean much more competition. But speaking as someone outside of Hollywood, that extra competition is exciting as it means a much wider variety of films will get made & the blockbusters of the future could be made by some kid in his bedroom.
enabling the blockbusters of the future being made by some kid in his bedroom
JJ shot on Super 8’s as a kid, Spielberg shot on his dad’s 8mm as a kid. Most directors had a camera and started off directing with their friends. This is also the route to YouTuber today, shooting videos as a kid with your friends. Where filmmakers develop their chops is in developing a coherent narrative, establishing characters and stakes that matter, capturing something that people aren’t expecting, or haven’t processed in a certain way before…etc.
I truly don’t think that AI is the inhibiting factor between an amateur film and a great film. That’s extremely reductive of the art. Maybe Sora will be some great bound and leap forward in storytelling capabilities, but making pretty visuals doesn’t make a good movie (Transformers 2 looked good, and was ass).
That being said, I’m quite anti-AI personally, so this is likely my biases bleeding through.
@mattyb put differently, i see 250-300 movies a year. the number that could’ve been improved by Sora is probably <10
I truly don’t think that AI is the inhibiting factor between an amateur film and a great film
You're right that it's not the factor, but it is a factor, as we don't talk about Spielberg's super8 home movies, we talk about Jaws, Schindler's List etc. which are studio films that cost a ton & took years to make. But ye, they're not solely good because they're pricey (otherwise Transformers 2 woudn't = 💩)
Since the dawn of cinema to present, the movies that have gotten made:
- take years to produce
- cost hundreds of thousands to millions (even if "low budget") to tens/hundreds of millions
- require a cast & crew & expensive equipment (spielberg may have learned with a Super8 but he didn't shoot E.T. or Jaws with it!)
- need to be produced to make a profit (which means even for a 'low budget' film at minimum they need tens of thousands of people willing to pay to see it)
- Need to be trusted with a huge amount of money & a deadline which requires a director with usually years/decades of experience
That means only a handful of people get to make movies & they're severly constrained (budget, time, need to make a profit) in the kinds of movies they can make & the number of opportunities they will ever get to do it
It's not so much about looking at the movies that get made today, but about the millions of movies that don't get made because they don't check all these boxes.
e.g. Imagine the next Spielberg gets to start making films in his teens that have the production value of a Jurassic Park & can be made in a few hours/days & he can distribute them to billions of people via the web. I suspect the stuff he makes will not be good at first, but he'll be able to hone his skills much faster & once he's good enough he'll only be constrained by the speed of his imagination & maybe how many FLOPs of compute he has 😄)
@mattyb 😂 He may not have a cellphone, but he's no luddite! His films do use CG when necessary.
Interestingly, he's on record as being optimistic about AI's use as a tool for filmmakers
Source: https://www.wired.com/story/christopher-nolan-oppenheimer-ai-apocalypse/
Interviewer: No, I don't think it does—especially when some of what we're talking about with AI is a softer threat. Disinformation on hyperspeed, technological unemployment.
Nolan: It is, but I'm less—I feel that AI can still be a very powerful tool for us. I'm optimistic about that. I really am. But we have to view it as a tool. The person who wields it still has to maintain responsibility for wielding that tool. If we accord AI the status of a human being, the way at some point legally we did with corporations, then yes, we're going to have huge problems.
Interviewer: Are you seeing anything in AI that could be wonderful for, in particular, filmmaking?
Nolan: Oh definitely. The whole machine learning as applied to deepfake technology, that's an extraordinary step forward in visual effects and in what you could do with audio. There will be wonderful things that will come out, longer term, in terms of environments, in terms of building a doorway or a window, in terms of pooling the massive data of what things look like, and how light reacts to materials. Those things are going to be enormously powerful tools.
Will you take advantage, personally?
I'm, you know, very much the old analog fusty filmmaker. I shoot on film. And I try to give the actors a complete reality around it. My position on technology as far as it relates to my work is that I want to use technology for what it's best for. Like if we do a stunt, a hazardous stunt. You could do it with much more visible wires, and then you just paint out the wires. Things like that.
It'll improve the ease and efficiency of visual effects, you're saying.
It's not starting from nothing. It's starting from a much more detailed and data-driven idea. It might finally break the barrier between animation and photography. Because it's a hybrid. If you tell an artist to, say, draw a picture of an astronaut, they're inventing from memory or looking at references. With AI, it's a different approach, where you're actually using the entire history of imagery.