Disney signed deal with OpenAI to bring its characters to AI-generated content. This includes Marvel and Star Wars.
There's going to be so much fucking porn, and Disney won't be able to do a thing about it since Sora works will be deemed acceptable license use. I sincerely hope this explodes in Disney's face so hard they fire everyone who made this decision and bring back people who have a functioning brain.
AI is only going to get better, and it's going to revolutionize films even where the characters aren't AI-generated. I think AI-generated films will become the norm. I think they'll be accepted (after a fashion) and in ten years or so, seen as necessary and inevitable. The idea of monstrous sound stages and sets built on back lots and costume departments and lighting departments and so on will be very quaint. Digital artists will have replaced all of that, and they'll be hypercharged with AI. But if I'm wrong, the market will correct. As always , if you don't like it, don't pay for it.
I think you're wrong. I think there will be a continuing pushback similar to what we've seen with that Tilly Norwood thing.
I'm gonna hang onto physical media as long as I can. I'll avoid AI movies for as long as I can, and if/when all movies get slopped, I'll read books, and if they slop up books, I'll re-read the old ones 'til I die, and be like the guy in "Fahrenheit 451" who lets himself burn with his library. The people that live in a slop-ified world and are happy about it will be like the zombie housewives in "Fahrenheit". Goodbye, real humans.
People in the business care about Tilly Norwood. Some in the audience care. Most schmoes aren't aware and wouldn't care if they were. They just want the "content" to keep flowing. But I think the content will improve with time. The massive productivity gains that AI will release will let individual creators become their own film studios.
I'm reminded of the scene in "The Dark Knight Strikes Again" where it's revealed President Rick Rickard is an AI hologram created by Lex Luthor. He jams and flickers on live TV, and one guy-on-the-street is like "who cares if he's not real!? He's a great American!!". Frank Miller's cheese has fallen off his cracker, but before that last little corner slipped off, he predicted this shit.
Those people wouldn't be "creators". Telling an algorithm to spit out an amalgamation of other people's stuff isn't creating anything .
I think you're selling people short, or assuming that AI is going to make all the creative decisions. I don't think that'll be the case. (Also, I'd point out that none of us is Adam in the morning; we all have our own influences. We all derive our designs with the awareness and knowledge of earlier creations. We're still writing dramas in a way Sophocles would readily recognize. A friend of mine likes to quip--ironically--that Shakespeare is just a bunch of old cliches.) Imagine you're a creator of limited means and you want to create a western drama, so you use AI to create a frontier town setting. You can add a saloon or a schoolhouse easily. Maybe you like the train station from High Noon, and the church from McCabe & Mrs. Miller. You can use those as a starting point and derive from them. It isn't going to be as easy as pressing a button and creative input will still be required, but after a few hours work, you've not only designed a town, but created all the virtual sets you need. I think this is going to unleash creativity, not hinder it. Imagine a single person or small team of people being able to produce a feature film that looks as good as one of today's big budget productions, but they did it from a small office. I could be wrong, and if I am, and AI just results in an endless stream of hot garbage then it deserves to fail. But I think it's not going to fail.
I think Star Trek got it right on this one saying that TV (at least in any way recognizable to a 20th century audience) was dead by 2040. It's long been a running theme that they are more into small scale live performances. Because when you can generate anything you want from the push of a button that generated material loses value. The valuable thing becomes seeing the act of the creative process. How you can share and shape an experience with others. People predicting a future where people are sitting down to watch AI generated blockbuster spectacles are probably right in what the technology will eventually allow, but not the world that will result from it. That view feels quaint, like a 19th century person speculating on a future where people can easily move around the country because it will be crisscrossed with canals where the boats are pulled by mechanical horses.
No offense, but you're speaking from the perspective of a non-creative. Every author filters stories through their personal experience. A story might have been told before, but never by you (Harry Crews). And if you are pulling the church from one TV show and the train station from another, you aren't creating anything. You aren't filtering things thrift your experiences. You're taking sunshine else's work, without their permission, and using it to make something for your own amusement.
Someone is guiding the creation of the material. That is the creative part. The guys who build sets and rig lighting are important, but there's little of their vision in the final product. And I wasn't proposing the outright theft of IP, but using it as a starting point...as creators often do. All of the conventions you see in film had their origin somewhere, and they became imitated. Even something as basic as a tracking shot had to be invented by someone. That's why films from the 1900s and 1910s look so flat and stiff...no one had figured out the language of cinema yet. There was plenty of idea "theft," but we don't look at it that way (and neither did early filmmakers). AI is going to streamline the production process; it isn't going to provide the vision for it. That part (speeding production) is definitely going to happen. That will make films easier to produce, perhaps enabling a single person to create one. Will it result in better films? I think so, though it will spawn a torrent of awful stuff as well. We'll see.
On a side note, if you haven't seen it, The Story of Film: An Odyssey is an excellent documentary miniseries from 2011. I encourage people in my line of work to just watch the first couple of episodes if they don't want to commit to the whole 15 parts because it talks about things like how revolutionary it was the first time someone made an edit. Something we do every day with a keystroke blew people's minds in the early 1900s. It originally ran on TCM and has bounced around different streamers. Currently I'm rewatching it on Prime.
Seen it and have the book. Terrific series. I read Walter Murch's book on film editing some time back and was quite astonished at how much psychological study has gone into it. One thing he wrote that really stuck with me: the edit often comes where you would blink. Your attention is ready to shift. In a film class I took once, we spent 15 minutes discussing an edit in Casablanca, where Ingrid Bergman gets up to leave (2:25 in this video). It is a terrific cut because it's practically invisible.
A match cut. Also known as sequencing. The edit happens in the middle of her movement. Not invisible, but if done properly...seamless. Makes it look like there were two cameras there when there was only one. I discovered match-cutting on my own at my first job and thought it was the coolest thing ever (even though I'd crossed the shooting axis while doing it). Mark Cousins's Orson Welles documentary is good, too. I recently rewatched that. Back when I posted a whole lot more on Twitter, Cousins and I had some great conversations. I keep that Murch book in a magazine rack in the downstairs bathroom. I frequently re-read parts of it while I'm in there.
I want to take the time to really highlight this, which is where much of the error in certain thought processes comes from. Yes, we are all influenced by the stories we've read. But as humans, we filter everything through our individual experiences. Each of us can read Romeo and Juliet and come away with different perspectives on the piece. AI doesn't do that. AI is incapable of that. AI cannot decide what it will read. It can't create its own opinion on a piece. It has no perspective it brings with it to the piece. All it can do is mathematically determine how simmering might turn out based on parameters it is given. AI is not "inspired" by anything. If the day comes when an AI can say "No, I don't want to read that, I hate Shakespeare.", then we can consider things differently. Until then, it is incapable of actual creativity.
True, a human being can say "I once attended a wedding at a little seaside church and that inspired me to create this design" and that is creative. AI can't do that. But an AI prompted to generate images of a small seaside church can draw from thousands of sources and quickly synthesize hundreds of varying images incorporating aspects of all of them. A human cannot do that. And there is still a human creator writing the prompt and evaluating the output. And if you're a creator and need a small seaside church for a scene in your movie, the 5000% increase in productivity that comes from typing a specific prompt over hiring an artist may justify the loss of personal attention that viewers will probably not even notice. If the output is indistinguishable to the viewer, there isn't much practical difference.
Can you not agree that there is something inarguably dystopian about outsourcing our culture to an artificial algorithm? Even if it is indistinguishable?
I don't see it as dystopian, though I acknowledge something is lost. But a lot is gained. And I don't see the algorithm controlling the culture. I see AI unleashing tremendous productivity. It will replace a lot of creative work with soulless calculation, yes, but it will enable the small creator to do large things.
I'm not sure what is gained beyond this nebulous idea that "small creators can do large things". I'm not even sure what you mean by that line anyway. But regardless of it, our art is what defines us as a culture. If a computer does that for us, how long until we just let computers do our thinking for us too?
Imagine a person who writes a story (maybe assisted by AI) and generates a feature film from it. The AI helps refine the dialog. The AI finds a plot hole the writer missed. All the sets, characters, etc. are AI-generated. The creator films a real actor and prompts the AI to use the performance. The AI generates a musical score. It's endless. Already, AI is letting non-artists create art. Already, companies are making commercials with it. Already, Disney has invested heavily in it. Crude as the technology still is, it shows astounding potential. Mmm, lots of things define our culture (science, language, morality, history, religion, philosophy, tradition, economy, humor, etc.). And many more people enjoy art than create it. If an AI generated a catchy song, would people refuse to listen to it because it was artificial? Some, I'm sure, but probably not most and certainly not all. If, say, an AI turned out to be significantly better than a typical doctor at finding cancer early in x-rays, would you object to using it? No, I'm sure. Sometimes, losing the human element is worth it.
I did an immersive AI experience in Las Vegas a while back called "Arc". It takes place in the future where humans have travelled to a distant planet. You basically walk around the space station orbiting the new homeworld and explore things. The graphics were okay (there's definitely better out there) but the mechanics were amazing, especially the feeling of artificial gravity. That's the future of entertainment, I think. Not sitting in a theatre and passively watching a movie.
I think the future isn't either/or. It's full spectrum. From dramas enacted on a bare stage to full immersion inside an interactive, story-driven video game.
Speaking strictly about the visuals: Having dabbled in AI a bit for fun*, I can say that idea-wise, it's only as good as the prompts you give it. Over the weekend I asked ChatGPT to make me an image and it took four or five tries before it gave me what I was really asking for. And this was a really simple image. Tech-wise, AI has a very long way to go IMHO before it's as realistic-looking as film and video footage or what top CGI artists can turn out. I suspect it'll eventually get there, but right now it's really easy to tell the difference. But even if/when AI gets good enough that you can't tell that it's AI, as with any kind of medium, the technology won't be able to save you if your story sucks in the first place and, as Paladin says, the market will let you know that your work stinks. *The only times I've use AI professionally are the one time I asked ChatGPT to gin up a disciplinary letter for one of my employees so it sounded more business-like and I frequently use Adobe Enhance Speech to clean up audio that has too much background or wind noise. That program is so crazy good that I usually only have it set to about 20% or so because, beyond that, the audio sounds fake and unrealistic. Because AI is, by definition, not real, I'd never use any kind of AI for journalism video work beyond possibly employing some sort of color-correcting thing to speed up the grading process (assuming such a program existed). Anything else would be journalistically unethical. I won't feel threatened by AI until they develop Camera-nators because AI literally can't do my job.
I agree it has a ways to go. But the "Will Smith eating spaghetti" meme shows how quickly it's advancing, and, with all the money being poured into it, I expect some big steps. I use AI (grok, Copilot, and Gemini/Sonnet/Claude) frequently for research. I've also used grok to critique ideas and tighten up paragraphs I've written for papers, and it does a pretty damned good job. Yes, AIs hallucinate, so caveat emptor. But everyday on my social media feed I'm seeing videos that I have to watch for a few seconds to conclude they're AI. I've seen fan-made Star Wars and Star Trek stuff that looks great and the characters even sound like the actors. Of course, the line delivery is clunky and the characters are stiff, but it's early days. We couldn't do this at all a couple of years ago. And I agree: prompt-writing is definitely a skill to cultivate if you want good results.
This is my biggest problem with AI. I also see photos and videos like that with people trying to pass them off as real ("Look at this plane crash!"). The Venn diagram between the chemtrail/flat earth/QAnon types and the people who can't tell that stuff is fake is a circle. AI imagery used for purely fiction? Sure. Knock yourself out if you want. AI imagery being misrepresented as a real event? Go fuck yourself.
