In Hollywood there seems to be two camps when it comes to AI.
Some preach how it will change how movies are made, but with few concrete examples to back up the capacity of the new technology. On the other side are anti-ai-zealotes, which contribute one oversized setback Over anyone who uses AI tools in their work.
To go bravely into the huge void between both is a writer/director Scott Z. BurnsAn AI skeptic who wanted to see if it was actually a tool he could use. Specifically, Burns wanted to see about large language models (LLM) that chatgpt could help him with a script for the sequel to his and director Steven SoderbergS “Infection“A trip he documented in the new Audible Original Podcast series” What can go wrong? ”
So is LLMS a useful screenwriter tool? Can you ever write a script yourself? Can it help to come up with a good idea for a movie? The answer is complicated, and the series with eight episodes does a good job of exploring the question from several perspectives, but overall Burns has passed away from the experiment which is much less concerned that AI will ever be a successful screenwriter.
“It can’t write scenes,” Burns said when he was a guest in an upcoming episode of Indieview Filmmaker tolkit podcast.
In a section of “What can go wrong?”Infection“And” lexer “, Burns’ Gen-AI collaborative partner. This is where the limitations of lexes, a custom chatgpt, are obvious. Regardless of the fast and modifications they feed lexes, it cannot capture the internal emotions and interpersonal dynamics between the two former colleagues.
“It has run out of gasoline when you start getting into some of the more detailed parts of filmmaking,” Burns said. “It does not have an experience of itself in three -dimensional space, and we, I think, underestimate what our physics makes available to us.”
However, writing scenes was not Burn’s goal to go into the experiment. “What I was interested in using it for was very quick to come up with a cool idea for a movie.”
“Contagion” was a hit when he was released in 2011 but found a completely different life in 2020 during the Covid -Lockdown, when millions were drawn to watch or see the thriller that caught a dramatized version of what we all lived through. To make the original, Burns and Soderbergh worked with several top epidemiologists, who helped get many details about how a modern pandemic would play, nine years before Covid met.
This success led to the desire to make a sequel, but the prerequisite for a second fictional but science -based pandemic, which came after the audience had experienced a real one, had been difficult for Burns and Soderbergh. Burns hoped that AI could quickly help explore several different science -based premises and help it quickly to it ideas based on “what about” issues until they found the seed from a premise that they could build up.
“AI is useful in that respect,” Burns said. “So if you have an idea and you are trying to go,” Ok, then what do we do with this idea? What are the permutations? What are possible things? “It’s really good at making lists.”
Burns and Soderbergh made a major breakthrough with lexes. Burns has requested that IndieWire not be too detailed so as not to destroy the podcast, but emphasized that this breakthrough came through a series of mistakes.
“Lexter history to me is what ends up being the point of this whole experiment,” Burns said. “Which is Lexter was not meant to be my writing partner, companion, whatever we want to call it; lexes were meant to be one (film) critics. And why we created a critic was on an opportunity that we had a conversation, and we thought, “Well, let’s ask a critic about a sequel to” infectiousness “is even a good idea.” So of course we had to create an AI critic, and it was lexes. ”
The sharp -heavy, British accentuated Siskel and Ebert Bot’s biggest contribution to the new condition was also a mistake or a happy accident. The idea for the prerequisite came from Lexter disregarding an important aspect of Burn’s fast and created a scenario outside the perimeter it was given.
Burns believes that the ability to quickly work through various scientific permutations of what the next pandemic can be, and therefore quickly offer a number of potential narrative paths to go down, play to Gen AIS strength. But can this type of brainstorming work for other movies? Burns is very hesitant, and his experience of lexes makes him believe that if Studios relies on AI to generate movie premises, they will end up with derivative work at best.
“We spent a lot of time in the podcast and wondered if the technology we used would actually work for other films,” Burns said. “And then suddenly you are,“ oh, so do I do what a streamer can do? “Reassemble the constituent parts of other films for a new movie.Aor. “”
And how do Burn’s screenwriter who use lexes look forward? He is currently running a writer’s room for a new Netflix series, and he sees how scared writers use AI.
“Although I sit here today, I have a lot of anxiety about how people will react to the fact that I have done this thought experiment,” Burns said. “Because there are many people in my guild, of which some I am very close friends with, who think AI is a violation.”
Burns believes that when screenwriters realize that AI cannot come up with original ideas, nor write scenes, they may be more open to using it.
“I encourage people not to be afraid. It’s not to write the scenes,” Burns said. “What it can do is if you have a character and you can fill the prompt with unique, very specific details about who that person is and what they want in the world, and then you say: ‘That person has to take their daughter on vacation because they try to join. Where should they go, and what should they do?” It will spit out five ideas really fast.
All eight episodes of “What can go wrong?” Are available now on audible and wherever you get your podcasts.
To make sure you don’t miss Scott Z. Burn’s upcoming episode of Filmmaker Tolkkit, subscribe to the podcast on AppleThe SpotifyOr your favorite podcast platform.
