A YouTuber turned tech executive has launched an AI media lab that’s betting on interactive video as the future of entertainment.
Ben Rellis, who turned his career as an early YouTube creator into an executive role at the video platform, is launching Make Believe, an artificial intelligence lab that will focus on creating technology that enables videos that can talk to viewers, betting that it will enable forms of entertainment that were never possible before.
“A lot of the talk in Hollywood is about how AI is going to make things cheaper and faster for film and TV, and I would say for us we’re focusing more on formats that were impossible before AI existed,” Rellis says. Hollywood Reporter In an interview.
Imagine, for example, a cooking video in which you can ask the chef questions while preparing a dish at home, or a fitness builder who can critique your figure, or a guitar instructor who helps guide you while teaching you a song.
“Some of this we can do already, some of this we need to build the tools to be able to do that, but I would say the general idea that platforms like YouTube have been a great place for people to learn to do new things, I think that can really be multiplied by what interactive video enables when video can adapt to you, and can see what you’re doing,” he explains. “For over 100 years, video has been something you watch. We believe part of the next era of video is something truly interactive.”
The Make Believe leadership team also includes fellow YouTube veteran Margaret Burris and Alec Lindsay, formerly of HeyGen. But Relles, as it happens, knows a thing or two in particular about what it takes to be creative.
He co-founded the Vsauce network and the Key of Awesome YouTube channels (you may remember their 2007 viral hit “Crush On Obama,” which has nearly 28 million views), with a total of seven billion views between them. He eventually joined YouTube, where he led innovation and unscripted programming at YouTube Originals.
While at YouTube, Rellis connected with Reid Hoffman, the co-founder of LinkedIn and an AI enthusiast. Hoffman is an investor in Make Believe, where the team’s technology is helping create “Reid AI,” an artificial intelligence avatar of Hoffman that is trained on his writings and speeches.
THRin an unusual arrangement, was referred to Reed Hoffman’s AI avatar for quotes about his decision to invest in Make Believe. “This is about amplifying human agency and communication through AI, not replacing it,” Avatar stated, without irony, of Hoffman’s investment.
“I think a big part of what this technology can do is give you access to people that you wouldn’t otherwise be able to reach,” Rellis says of the technology his company is creating. “And the next step was really creating the best real-time video avatar, an avatar that you can talk to and have a conversation with and answer in real time, and that can leverage the right interview or book or podcast, depending on your question, and so we started finding a lot of different use cases for this real-time avatar.”
Make Believe already has a deal with A+E Networks’ History Channel to develop technology for use in its wheelhouse.
“People learn visually, and they learn through interaction,” Rellis says. “I see a lot of applications through science and math.” “Thinking about how we learn about history by having conversations with historical figures about the times they lived firsthand, I would say that’s a general category that I think is going to be a space that we want to try and build, for sure.”
But AI video’s current moment reminds Relise of those early days of YouTube, when it was clear that a new form of entertainment was emerging, but its ultimate form wasn’t quite clear yet.
“I always loved the first few years of YouTube, where these new formats had a breakthrough that no one expected: unboxing toys, hair tutorials, playing with toys, and I think it will be the same here,” he says. “I have real convictions about where that will be most valuable, but I also realize that the specific formats that will be most successful in interactive content will require a lot of experimentation and trial and error.”
