I finally watched Rogue One this week and maybe the most remarkable thing about that movie is that one of the main characters, Grand Moff Tarkin, is played by someone who died in 1994. This wasn’t simply a case of highly resourceful film editors repurposing unused footage from other films – he was recreated using cutting-edge CGI.

2016 has been a rough year for celebrities… it seems like almost every day another beloved icon is added to the long list of “people we lost this year.” Simply being dead is not going to stop these celebrities from starring in the future’s biggest blockbusters, though.

We’ve seen CGI versions of stars before of course. A good example is in Tron: Legacy, where filmmakers needed to “turn back the clock” and recreate the 1982 version of Jeff Bridges from the original masterpiece, Tron. This was fairly convincing as long as you didn’t look too closely or notice his mouth, which just looked weird. However, that was six years ago, and the technology and tools have been rapidly advancing and improving. I watched it Rogue One on a giant IMAX screen in 3D, and even in that format you could be forgiven for not noticing that Grand Moff Tarkin wasn’t real – it will be hardly noticeable when people are streaming at home on their relatively tiny 65″ screen. If Rogue One is any indication, it won’t be long until CGI actors are indistinguishable from their living co-stars.

There are so many reasons that we’re likely to see a lot more of digitally revived stars such as Cary Grant, James Dean, Marilyn Monroe, Robin Williams, Elvis Presley, Elizabeth Taylor, and Frank Sinatra in movies in the next few years. Even many movies with younger versions of still-living actors will be common.

To keep this “Bite”-sized here are just a few:

  • The technology will be able to do it convincingly – within a few years, I doubt people will be able to tell who’s real and who’s CGI’d (unless they know that the actor is dead)
  • CGI stars will likely be MUCH less expensive than living actors, especially if they’re completely fictional and not based on a real person
  • CGI stars won’t have crazy demands of the studios, and they’ll never delay filming because they’re late to the set or drunk or high, etc.
  • They also won’t have scheduling conflicts with other projects. Countless movies have starred people other than who the part was written for because of scheduling conflicts
  • CGI stars can be tweaked for each role. Christain Bale had to practically starve himself to death for his role in The Machinist. That won’t be necessary anymore
  • No restrictions on the amount of time that CGI stars can “work,” this will be especially important when replacing child actors who have strict limitations on their schedules
  • Reshoots won’t be a problem, even if the original filming was done years ago
  • Living actors can begin licensing their images so that their CGI versions could work on multiple projects at the same time. If the tech is good enough, actors can decide that they only want to work on projects that they’re interested in and audiences will never know if it’s actually them or their CGI clones
  • Entirely new stars can be created based on viewer preferences (much like how Netflix creates original programming based on subscriber usage patterns). Studios could “own” their CGI stars, similar to how it was in the early years of filmmaking when an actor was in a contract with single studio

Finally, the real breakthrough will be when CGI actors can do their own press tours live. The next question is, at what point does using a camera to make a movie become like shooting a movie in B&W today – just a throwback to another era of technical limitations.

BONUS BITE: To take this a few steps further, it’s not impossible that within the next five years we will have movies created for us on demand. You could say “I’m looking for a comedy like Anchorman, set in Alaska in the 1990’s, starring Richard Pryor and Carol Burnett” and software would create that movie just for you. It would use all available data about you, including your current mood, what you had for breakfast, anything you’ve written, your reactions to various parts of other movies –  anything and everything will be used to create a hyper-personalized movie that you would enjoy the most.

Later, advanced versions will be able to create movies based on the data from more than one person, so couples, families, or groups of friends can enjoy a movie together (maybe even a “movie experience” in VR.) It sounds crazy, but it’s probably not that far off. Hyper-personalized music creation will be first though. But that’s topic for another “Bite.”