How we made a movie by an AI screenwriter
The scenario starts off quite simply: A couple is at the end of dinner. The conversation ends, the wine almost finished. After a silence, the man says he wants to play a game.
Enter artificial intelligence.
Does AI dream of writing scripts?
Using GPT-3, we developed a short film script called A romantic date. Tired of offbeat AI like Cleverbot, we wanted to use more robust technology in our work. In our experience, previous iterations of AI were hijacked, easily confused, and lacked significant memory. You can tell Cleverbot your name, but the program can forget it in all four entries. Sometimes he immediately forgets. We wanted to find something with the ability to be self-referential. We needed something with the memory capacity.
While looking for a better solution, we came across Shortly AI. The site is aimed at people suffering from writer’s block, encouraging writers to overcome their frustration by using artificial intelligence. The program adapts to your work.
This means that if you write a western it will attract cowboys. If you write a space adventure, it will keep the atmosphere going. If you are writing a horror story, it will do its best to scare you. Of course, most of the time it deteriorates further, but for a while it can keep the tone of a story going. Impressed, we quickly tried to diversify.
Could he create song lyrics? Yes. Receipts? Yes. So what about a movie script?
To our delight, it could. As long as we captured the beginnings of a script, it followed a script format, including stage titles, dialogue, and action lines. We generated a few and decided to pick one to make.
Build a script
The entry for A romantic date Washingtonas following :
INT. BENNY HOUSE – NIGHT
BENNY and JULIA (ALL TWO 20 YEARS OLD) are sitting at both ends of a long table, drinking wine.
BENNY: I want to play a li‘I’m playing. It is a li‘I’m crazy, it’s a li‘I’m having fun, but it will blow your mind.
We generated three variations of the script. In the first, Benny accuses Julia of being pregnant and not loving her and then stabs her in the stomach. In the second, dinner becomes a paranormal mystery when Benny goes missing, leading Julia to search for him. In the version we ended up filming, Benny hypnotizes Julia and starts dancing on the table. When he jumps, he controls his body and prevents it from moving. She escapes slapping him and mumbles, “I felt weird.”
At this point, the AI ââhas struggled to stay on track. The characters often reorient themselves towards new oscillating emotions and live in a sort of heightened melodramatic reality. To the cinema‘At the end, they even talk about the movie itself, referring to previous fictional events.
In the lines of action, the AI ââindicated: â(An imminent break) JULIA: ‘I think we are in a movie.‘“
“Remind me, what is this?” Benny asks, referring to the very movie he’s in. “He hypnotizes her,” said Julia. ” It’s dark.
Sometimes the AI ââeven made some typos. For example, when Benny should say “Only now, “he wrote,” No now. “This gave a touch of realism to the proceedings.
Satisfied with the AI’s work, we decided to bring our messy and frantic script to the screen.
Bring the computer vision to life
Through Calamity AI, we aim to showcase the results of AI and humans working in tandem. We want our films to capture both the surprising accomplishments and the gaps in the collaboration. A romantic date was a great case study either way.
The limitations of artificial intelligence prevent it from doing every part of the filmmaking process. For example, GPT-3 can‘t develop their own list of shots or use a camera to film the project. Currently, a human is required to bring a piece to life on screen.
We believe this will be the future of technological creativity in cinema for a while. When a writer is in trouble, he can use AI to move his scripts forward. The screenwriter will still have to provide the initial input and then decide which ideas are good and bad. With our work, we hope to display the kind of content, however rudimentary, that a human can create with the help of AI.
As such, our first and foremost rule is that we take scripts seriously. We approach these scripts as if we are directors for hire, trying to make the best possible film out of what we are given. From cinematography to directing through the actors, we hope to approach the projects with seriousness.
For this specific film, we called on our friend Henry Chastain as director of photography. He’s good at mimicking styles from movies. We figured the story could be especially edgy if it looked naturalistic, taking place in a normal-looking house with normal lighting. We also chose to shoot using objective camera angles. We wanted the camera to look at the characters as if the lens was really looking at real people. Jean Cassavettes A woman under the influence was our visual reference.
We had worked with the actors, Owen Painter and Lily Rohren, on previous projects and knew they were the ones for the job. While they can both be comedic, both of their styles of acting are rooted in drama. Both actors are particularly good at playing real people. It was essential for the seriousness of the project. If they despised the machine-written lines, perhaps they delivered them in a goofy or stilted style, thwarting our philosophy of taking the script seriously.
As ridiculous as the script on paper was, we wanted to see if the AI ââcould provide the framework for actors to develop some really interesting characters. We told Owen and Lily to imagine the script was written by a real person, and they were both immediately willing and able.
We had a Zoom rehearsal before the shoot to discuss the material. We didn’t want the performances to be ironic. Achieving this type of sincerity can be difficult, however, when one of the first artificially written lines is Owen insisting that Lily’s spirit animal is a “dog.”
In a few takes, however, we had found the characters. Benny seemed genuinely convinced of his power, terrifying Julia. She communicated the terror of suddenly finding herself unable to control her body, having taken it over at what she thought was a romantic dinner.
Making of the movie
For the actual shoot, we had Henry on camera and ourselves on sound. It was a small-scale crew with each playing different roles. We also kept a close eye on the script; Owen and Lily had to be perfect to ensure AI writing fidelity. Owen put everything in the performance. He finished each shot on the table panting, his hands on his knees. The AI ââwould have appreciated his dedication.
But even artificial assistance cannot stop human error. With more videos slated for earlier release dates, we’ve been pushing A romantic date next to. We categorized the images and didn’t realize that we had failed to save the audio. Two months later, at the start of the editing, the audio SD card had been erased.
Fortunately, Henry’s camera picked up the source audio, but it wasn’t perfect. With a little bit of cleaning and tune-up work, we were able to get something that worked, although the performers often seemed to go from whispering to screaming for no rhyme or reason. This product wasn’t ideal, as a few eagle-eared listeners on YouTube and Reddit pointed out, but we got away with it.
The editing process offered many avenues to follow A romantic date: sinister, calm, bombastic, romantic. We landed on a hybrid approach, a melodramatic combination that was both romance and horror.
With A romantic date finished, the time had come to release him. For us, producing high quality work is always first. Audience metrics are somewhat beyond our control, but we can be masters of the quality of our material. But we have a publishing strategy that starts with consistency. We’re releasing a new video every other Tuesday, trying to keep the momentum going by attracting new subscribers and viewers.
For each video, we alternate between two thumbnails, testing which one gets the most attention. We post the video and encourage our friends to share it on social media, text it to family, and also share it on applicable subreddits. As views and comments start to flow in, we do our best to respond and be transparent about the process that went into the job. With this film, we were happy to attract the attention of various media.
AI: Coming to a theater near you?
Although we’ve been making movies together for many years, we never expected our third collaborator to be an AI. For a long time, we just kept making videos for ourselves to improve ourselves as filmmakers and hone our skills. This focus allowed us to be ready when this opportunity arose.
When we found out about the AI ââprogram and tried to write scripts with it, we knew we had to run it. Making the movies in the highest quality possible helps viewers more easily enter the absurd world that AI has created. Presenting a cohesive vision will allow people to see the potential here.
Now we want to take that same script and pass it on. We wonder how another filmmaker could interpret it. Imagine the actors are middle aged and near the end of a strained marriage. Or that the only background noise is a record player right next to the needle. Lights may be dim or the table may be in the center of a crowded restaurant. The characters could be drunk or maybe on their first date. Any of these changes would result in an entirely different experience for the viewer.
All of these things will be influenced by the director’s interpretation of the script. Despite all the technological advances in natural language processing, interpretation still limits AI’s control over the finished product.
To bring a machine-written script to life, artificial intelligence must work in tandem with a living person. Right now, the AI ââcan provide the script. We can probably train him to produce a shot-list or even a musical accompaniment. Ultimately, however, the end product must be interpreted and made by a human. For now, at least.