Miniature Rivian Trucks, AI and where I think media is going.

A bit of background on the process of making a stop motion film for Rivian and where I think AI will take media.

The process of making Gear Guard started with a few phone calls with @3rd_larry in the late summer of 2020. The idea for the film was to make a stop-motion piece showing the humanity of the security system in a way that was unexpected for a startup electric car company. @andrewaller and I moved fast from there, and by Christmas, we had storyboards, concept art for sets, puppets, and CAD files for the R1T. We started working on the puppet, and @circa1964 started building sets and fabricating two versions of the R1T, one in 1/6 scale and one in 1/36 scale. Shelby built the driving sets and assembled the larger set, and Matt and Jesse figured out how we were going to actually shoot the thing. 

Making something of this scale during the height of COVID was challenging. Animators in Portland were busy working on Windel, Wild, and Pinocchio, and we struggled to find a good match. During the process of building the sets and puppets, the roof of the R1T changed, and then after we finished updating the truck, we had to redesign the look of the puppet, changing from a silicone-based puppet to a fur puppet. Finally, in February 2022, Matt fired the last frame and photography for Gear Guard wrapped. 

Then the piece sat. After wrapping up GG and a few commercials for Nike, I took a break from filmmaking to gather my thoughts. My ability as a creative person isn't in my raw visual acumen or skill at managing a project, but in hearing the proverbial jingle of the ice cream truck before the other kids on the block and meeting it at the stop sign. I started @moviemountain in 2015 because I realized that everything on social media was already made up, so why not have some fun with it? Stop-motion requires a lot of work to build the sets, puppets, and animation, as well as a lot of proprietary gear. The result is a unique look that's hard to replicate. We have our niche and have made short films and commercials, and we are working on a feature film now. 

In the fall of 2022, I started seeing images from Mid-Journey that were indistinguishable from reality. At first, I had no idea what I was looking at. Images that were so elaborate and well referenced that I assumed they were from some cult film that, until now, had passed me by. With shock that I still remember a year and a half later, I slowly realized that these were created by some AI engine, with the only human input being a list of prompts and references.

Where will this go from here?  Well, give it 5–10 years, and AI engines will be able to crank out feature films for orders of magnitude less money than traditional productions. Production Studios value will be in the archives of IP and old films that the engines will be able to reference when creating their prequels, spin-offs, and sequels. AI engines will be able to pull actors from any period of their careers and make films that are indistinguishable from reality. Cameras will largely be a thing of the past, with composite images from your phone or film archives allowing composites of every person to be made. Beautiful images and videos will cost next to nothing to make, and people will become numb to it all. Commercial photography will be a thing of the past, with celebrities licensing their image to companies and then approving the final assets without ever setting foot on set. Advertising Agencies will be able to conjure the perfectly inclusive cast of body positive families hiking the Pacific Crest Trail and practicing yoga in the morning to sell Wellbutrin. As technology advances, an AI algorithm will spit out “unique” films and commercials, all with a similar message but made specifically for our taste, similar to an Instagram Explore Page today.

Most filmmakers I know, myself included, are sustained by making commercials. Companies and ad agencies don't pay artists to make commercials out of compassion but because no cheaper alternatives exist. If and when an alternative presents itself that makes a product that's indistinguishable from the old way but at a fraction of the cost, they'll do that nine times out of 10. In the two years since we finished working on this film, I've come to the realization that there's a good chance that this will be the last of its kind, the highwater mark before AI. Hell, maybe I'm wrong, but if not, the experience of making something by hand with a group of people will be the thing I miss.