Disney creates task force to explore AI and cut costs: Sources
As evidence of its interest, Disney has 11 current job openings seeking candidates with expertise in artificial intelligence or machine learning
Walt Disney (DIS.N) has created a task force to study artificial intelligence and how it can be applied across the entertainment conglomerate, even as Hollywood writers and actors battle to limit the industry's exploitation of the technology.
Launched earlier this year, before the Hollywood writers' strike, the group is looking to develop AI applications in-house as well as form partnerships with startups, three sources told Reuters.
As evidence of its interest, Disney has 11 current job openings seeking candidates with expertise in artificial intelligence or machine learning.
The positions touch virtually every corner of the company - from Walt Disney Studios to the company's theme parks and engineering group, Walt Disney Imagineering, to Disney-branded television and the advertising team, which is looking to build a "next-generation" AI-powered ad system, according to the job ad descriptions.
Keep updated, follow The Business Standard's Google news channel
A Disney spokesperson declined to comment.
One of the sources, an internal advocate who spoke on condition of anonymity because of the sensitivity of the subject, said legacy media companies like Disney must either figure out AI or risk obsolescence.
This supporter sees AI as one tool to help control the soaring costs of movie and television production, which can swell to $300 million for a major film release like "Indiana Jones and the Dial of Destiny" or "The Little Mermaid." Such budgets require equally massive box office returns simply to break even. Cost savings would be realized over time, the person said.
For its parks business, AI could enhance customer support or create novel interactions, said the second source as well as a former Disney Imagineer, who declined to be identified because he was not authorised to speak publicly.
The former Imagineer pointed to Project Kiwi, which used machine-learning techniques to create Baby Groot, a small, free-roaming robot that mimics the "Guardians of the Galaxy" character's movements and personality.
Machine learning, the branch of AI that gives computers the ability to learn without being programmed, informs its vision systems, so it is able to recognize and navigate objects in its environment. Someday, Baby Groot will interact with guests, the former Imagineer said.
AI has become a powder keg in Hollywood, where writers and actors view it as an existential threat to jobs. It is a central issue in contract negotiations with the Screen Actors Guild and the Writers Guild of America, both of which are on strike.
Disney has been careful about how it discusses AI in public. The visual effects supervisors who worked on the latest "Indiana Jones" movie emphasized the painstaking labors of more than 100 artists who spent three years seeking to "de-age" Harrison Ford so that the octogenarian actor could appear as his younger self in the early minutes of the film.
'STEAMBOAT WILLIE'
Disney has invested in technological innovation since its earliest days. In 1928 it debuted "Steamboat Willie", the first cartoon to feature a synchronized soundtrack. It now holds more than 4,000 patents with applications in theme parks, films and merchandise, according to a search of the U.S. Patent and Trademark Office records.
Bob Iger, now in his second stint as Disney's chief executive, made the embrace of technology one of his three priorities when he was first named CEO in 2005.
Three years later, the company announced a major research and development initiative with top technology universities around the world, funding labs at the Swiss Federal Institute of Technology in Zurich and Carnegie Mellon University in Pittsburgh, Pennsylvania. It closed the Pittsburgh lab in 2018.
Disney's US research group has developed a mixed-reality technology called "Magic Bench" that allows people to share a space with a virtual character on screen, without need for special glasses.
In Switzerland, Disney Research has been exploring AI, machine learning and visual computing, according to its website. It has spent the last decade creating "digital humans" that it describes as "indistinguishable" from their corporeal counterparts, or fantasy characters "puppeteered" by actors.
This technology is used to augment digital effects, not replace human actors, according to a source familiar with the matter.
Its Medusa performance capture system has been used to reconstruct actors' faces without using traditional motion-capture techniques, and this technology has been used in more than 40 films, including Marvel Entertainment's "Black Panther: Wakanda Forever."
"AI research at Disney goes back a very long time and revolves around all the things you see being discussed today: Can we have something that helps us make movies, games, or conversational robots inside theme parks that people can talk to?" said one executive who has worked with Disney.
Hao Li, CEO and co-founder of Pinscreen, a Los Angeles-based company that creates AI-driven virtual avatars, said he worked on multiple research papers with Disney's lab while studying in Zurich from 2006 to 2010.
"They basically do research on anything based on performance capture of humans, creating digital faces," said Li, a former research lead at Disney-owned Industrial Light & Magic. "Some of these techniques will be adopted by Disney entities."
Disney Imagineering last year unveiled the company's first initiatives in an AI-driven character experience, the D3-09 cabin droid in the Star Wars Galactic Starcruiser hotel, which answered questions on a video screen and learned and changed based on conversations with guests.
"Not only is she a great character to interact with and always available in your cabin, which I think is very cool, behind the scenes, it's a very cool piece of technology," Imagineering executive Scott Trowbridge said at the time.
Reporting by Dawn Chmielewski in Los Angeles and Krystal Hu in New York; Editing by Kenneth Li and Matthew Lewis
