as a Hollywood executive They argue that it is “totally unrealistic” to pay more to actors who spend extravagantly on AI programs (87% of those earning less than $26,000).
Entertainment companies like Disney have refused to go into detail about the nature of their artificial intelligence investments, but job postings and financial disclosures examined by The Intercept reveal new details about the extent of their adoption of artificial intelligence.
In one case, Netflix is offering up to $900,000 for a single AI product manager.
Unions of Hollywood actors and writers went on strike this summer for the first time since 1960, demanding better wages and regulation of studios’ use of artificial intelligence.
Shortly after the actors’ strike was approved, the Motion Picture and Television Producers Alliance (a trade group representing television and film companies in negotiations with the Actors and Writers Guild) announced a “groundbreaking AI proposal to protect digital likenesses of actors for SAG-AFTRA members.”
The offer prompted comparisons to an episode of the dystopian sci-fi TV series Black Mirror, which saw actress Salma Hayek trapped in a Kafka-esque struggle with a studio that was using her scanned digital portrait against her will.
“Having been poor and rich in this business, I can assure you that I have plenty of money at my disposal. It’s just a matter of priorities.”
“So $900,000 a year for each godless AI soldier on that income, when that income could afford SAG-AFTRA health insurance for 35 actors and their families, is just outrageous,” actor Rob Delaney, who starred in the “Black Mirror” episode, told The Intercept. “Having been poor and rich in this business, I can assure you that I have plenty of money at my disposal. It’s just a matter of priorities.”
Impressive actors’ demands include protections against AI manipulation of scanned portraits without adequate compensation for the actors.
“They propose that the background performers should be able to be scanned and receive a day’s pay, and that the scans, their images and likenesses should be owned by their company and can be used forever in whatever project they want without consent or compensation,” said Duncan Crabtree Ireland, chief negotiator for the actors’ union SAG AFTRA.
Entertainment writers also have to deal with replacing their work with AI programs like ChatGPT that can generate text in response to queries. Writers represented by the Writers Guild of America have been on strike since May 7 demanding work safety measures against AI. John August, a screenwriter for films such as “Big Fish” and “Charlie’s Angels,” explained that the WGA wanted to make sure “ChatGPT and his cousins cannot be credited with writing the script.”
Photo: Vuk Valcic/Sipa via AP Images
protect the portrait of an actor
Per the SAG-AFTRA contract, the daily rate for background actors can be around $200. A job posted by Realeyes offers a slightly higher price. Two hours of work he is $300 “express”.[ing] Various emotions” and “improvisation”[ing] To “train an AI database to better represent human emotions.”
Realeyes develops technology that measures user attention and reaction to video content. The post doesn’t mention any partnerships with streaming companies, but the videos on Realeyes’ website prominently feature Netflix and Hulu logos.
The post was specially prepared to attract strike workers, emphasizing that the gig was for “research” purposes and therefore “not a strike target”. “Keep in mind that this project is not intended to replace actors and requires the expertise of actors,” Realeyes said, stressing repeatedly that regulations could be circumvented by training AI to create “expressive avatars.”
“The ‘research’ side of this matter is primarily red herring. Industry research also applies to commercial products. ”
Experts question whether the line between research and commercial activity is really that clear. “If this ‘research’ is commercialized, it will almost certainly be used to build digital actors to replace humans,” said Ben Zhao, a computer science professor at the University of Chicago. “The ‘research’ side of this is primarily red herring,” he added, adding that “industry research extends to commercial products.”
“This is the same decoy switch that LAION and OpenAI pulled years ago,” Zhao said, referring to the Large-scale Artificial Intelligence Open Network, the German nonprofit that developed the AI chatbot OpenAssistant. OpenAI is a non-profit organization that created AI programs such as ChatGPT and DALL-E. “Everything on the internet can be downloaded, and because we do non-commercial research, we don’t have to worry about copyright. How convenient our commercial products are! ”
Netflix AI Manager
Netflix’s posting for a $900,000-a-year AI product manager reveals that AI is much more than just the algorithms that decide which shows to recommend to users.
The list notes the use of AI in content creation: “Artificial intelligence is driving innovation in all areas of business, including helping us ‘create great content. A post from Netflix’s AI product manager refers to the company’s “machine learning platform,” which includes AI experts “across Netflix,” and hints at a broader effort by the company to adopt AI. (Netflix did not immediately respond to a request for comment.)
The research section of Netflix’s website describes the machine learning platform, which has historically been used for things like recommendations, but now says it’s being applied to content creation. “Historically, personalization has been our best-known area, where machine learning powers our recommendation algorithms. We also use machine learning to build our catalog of movies and TV shows by learning the traits that make content successful.
Netflix is already leveraging AI technology. On July 6th, the streaming service premiered its new Spanish reality dating series, Deepfake Love. The series uses face and body scans of contestants to create AI-generated “deepfake” simulations of themselves.
In another job, Netflix is looking for a Technical Director of Generative AI for a game studio’s R&D tech lab. (Voice actors and writers are often used in video games.)
Generative AI is a type of AI that can generate text, images, and videos from input data. This is an important component of original content creation, but can also be used for other purposes such as advertising. Generative AI is different from the old and familiar AI models that provide things like algorithmic recommendations and genre tags.
“All these models are usually called discriminative models or classifiers. They tell us what something is,” Zhao explained. “We don’t generate content like ChatGPT or image generation models.”
“Generative models have ethical issues,” he said, explaining how classifiers are based on judicious use of limited training data, such as viewing history, to generate recommendations.
Netflix is offering up to $650,000 for the role of Generative AI Technical Director.
Video game creators have expressed concern about losing jobs to generative AI, with major game developer Ubisoft saying it is already using it to create dialogue for non-player characters.
Netflix touts one of its games, Scriptic: Crime Stories, a narrative-driven adventure game centered around crime tales, as “using generative AI to tell its story.”
Disney AI Operations
Disney also lists job openings for AI-related jobs. For one, the entertainment giant is looking for a senior AI engineer to “drive innovation across its film pipeline and theatrical experience.” The post mentions several big-name Disney studios where AI is already playing a role, including Marvel, Walt Disney Animation, and Pixar.
During a recent earnings call, Disney CEO Bob Iger hinted at the challenges the company will face in integrating AI into its current business model.
As journalist Li Fang recently reported, Iger said, “In fact, we are already starting to use AI to increase efficiency and ultimately serve our consumers better.” “But it is clear that AI is going to be very disruptive and can be very difficult to manage, especially from an intellectual property management perspective.”
“I can say that our legal team is already working overtime to address some of the challenges that may arise here,” Iger added. Iger didn’t go into specifics, but Disney’s Securities and Exchange Commission filings provide some clues.
“It is clear that the entertainment industry is willing to invest heavily in generative AI.”
“Rules governing new technological developments, such as the development of generative AI, remain unresolved, and these developments may impact aspects of our existing business model, such as our revenue streams from the use of our intellectual property and how we create our entertainment products,” the filing states.
Among the union demands, which Mr. Iger judged “unrealistic,” are the striking actors trying to protect their intellectual property from AI, as well as Disney.
“It is clear that the entertainment industry is willing to invest heavily in generative AI. Not only is it potentially worth hundreds of millions of dollars, but it also provides valuable access to intellectual property so that we can train AI models to replace human creatives such as actors, writers and journalists for a fraction of the wages of humans,” said Zhao.
For some actors, this is not a battle against the sci-fi dystopia of AI per se, but simply a bid for fair working conditions in their industry and control over their appearance, body, movements, and speech patterns.
“AI is not bad, it just requires workers (me) to own and control the means of production!” Delaney said. “My melodic voice? My broad shoulders and my dancer’s wavy ass? I’ll decide how to use it! Not a board meeting of VC angel investor bastards while doing something over a cocktail of niacin drips in a Sun Valley boardroom.”
