The Character Is Disposable. The Infrastructure Is Permanent.

Every time new synthetic character technology arrives, the same script plays out.

Critics call it an abomination. Purists say it will destroy their industry. The company behind the experiment either folds or retreats. Then, quietly, the technology becomes the backbone of everything.

We are living through this cycle right now with AI-generated digital performers. Tilly Norwood is an AI character created by Proxy Studios. She faces the same disgust and anxiety that greeted every previous wave of synthetic character technology.

Stop asking whether audiences will accept AI characters. History already answered that.

Ask instead what separates the experiments that reshape industries from the ones that get buried in the graveyard. Because the graveyard is real. It is full of companies with the right technology and the wrong strategy. Understanding the difference is worth millions in positioning, investment, and deal-making over the next five years.

TL;DR

Every wave of synthetic character technology follows the same arc: backlash, underground development, mass adoption.

The companies that have consistently lost confused the character with the technology, like Square which closed after Aki Ross flopped in 2001, even as its render pipeline quietly migrated into ILM and Weta.

The companies that won treated the character as a proof of concept for the infrastructure, like George Lucas building ILM, like Serkis building Imaginarium Studios.

Three things your business needs to know right now:

The Creator Operator is a reader-supported publication. If someone forwarded you this issue, subscribe at rodrigoabdalla.beehiiv.com.

In 1999, George Lucas introduced Jar Jar Binks, the first major CGI character designed as a fully integrated protagonist. Critics called him an embarrassment. Ahmed Best, the performer behind the character, has said the backlash nearly destroyed him personally.

Twenty-five years later, Jar Jar is a cult figure. Best returned to Star Wars at The Mandalorian to a standing ovation. The pipeline Lucas built at ILM became the foundation of every major visual effects studio on earth.

In 2002, Peter Jackson introduced Gollum. The Academy refused Andy Serkis an Oscar nomination. Twice! Serkis went on to build Imaginarium Studios and pioneered a performance discipline that now employs thousands.

The same critics, the same complaints, the same "this will destroy the industry" argument has arrived for Tilly Norwood.

The backlash-to-adoption pattern for Jar Jar, Gollum, and every major synthetic character technology are nearly identical. It is not coincidence but a reliable leading indicator that a technology has crossed from experiment to genuine threat, which historically happens five to seven years before mass adoption.

Read the current backlash as evidence of failure and you are reading the data backwards.

The fact that synthetic character technology always wins does not mean every company betting on it wins.

In 2001, Square released Final Fantasy: The Spirits Within, a fully photorealistic CGI film built around a synthetic human protagonist. Square designed Aki Ross to be the first digital movie star. She appeared on the cover of Maxim. The stated goal was that audiences would believe she was human.

That is the word-for-word goal of today's AI character studios, twenty-four years earlier.

The Spirits Within cost $137 million and grossed $85 million. Square Pictures closed. Aki Ross disappeared. But the technology did not fail. The story did. Critics praised the render pipeline even as they buried the film.

Within five years, that pipeline had migrated into Weta Digital and ILM. Every face in every Marvel film owes something to what Sakaguchi's team built for Aki Ross.

This is not a coincidence either.

In 2001, SAG's own spokesman predicted all of this. He said they could "hire an awful lot of our members for a lot less" than it cost to create Aki, then added that "once the price falls a lot, that might make the situation more disconcerting." Twenty-four years later, Tilly's CEO jokes about agents calling. The price fell to essentially zero.

The Aki Ross lesson is not that synthetic human characters fail. Square confused its character with its technology. The character is disposable. The infrastructure is permanent. The companies that quietly kept building the capability won.

James Cameron has spent thirty-five years proving what the right strategy looks like. His success came from smart sequencing, not ambition or budget.

He moved the audience one step further than they would refuse, and never further than that.

In The Abyss (1989), the CGI creature is an alien. Audiences have no template for what an alien looks like, so there is no uncanny valley to fall into. Cameron builds the pipeline on something audiences have no basis to reject.

In Terminator 2 (1991), the T-1000 looks human but is a machine. Any audience discomfort is justified by the story itself. Cameron extends the boundary without betting everything on a single leap.

In Avatar (2009), twenty years of compounding investment pays out. CGI protagonists are now possible. But Cameron still does not place a photorealistic human at the center of the story. He uses the same alien buffer he deployed in 1989.

Audience tolerance for synthetic characters builds in a specific order. Alien first. Then machine-mimicking-human. Then stylized-human. Then photorealistic-human. Cameron cashed a check he spent twenty years writing. Square tried to cash it without writing it first.

Van der Velden's decision to keep Tilly in the AI genre is closer to Cameron's logic than most critics acknowledge. She is finding her alien register, a space where audiences carry no prior expectations. The mistake is treating this as a permanent home rather than a deliberate on-ramp. Cameron never stayed in the alien genre. He used it to build toward something bigger.

The current conversation obsesses over institutional legitimacy. People ask if Tilly can get an agent. They wonder if she deserves SAG recognition. They question her awards eligibility.

Andy Serkis never got his Oscar nomination. He got something more valuable. Imaginarium Studios. A technique every major studio depends on. The institution lagged the market by fifteen years. Serkis built the market infrastructure while he waited. By the time the Academy was ready for the conversation, he had already won.

The SAG-AFTRA 2023 AI provisions reveal which war is actually worth winning.

The contract creates two categories.

Tilly Norwood, designed from scratch with no real human as the template, is a Synthetic Performer. That is a meaningful legal distinction.

SAG designed the provisions to protect human actors from likeness replication. They were not designed to prevent the creation of original AI characters. In building a moat around Hollywood productions, SAG inadvertently drew a map of exactly where AI character studios can operate freely: digital-native platforms, brand content, gaming, social media, and international distribution.

That open territory is enormous and essentially unguarded.

For creators, athletes, and celebrities reading this: the SAG provisions should be a template for your own contract language, not a distant Hollywood story. The digital replica protections that SAG members now hold are protections most creators do not currently have for their own likeness. If your management has not updated your standard agreements to include the following, you are less protected than a Hollywood background actor.

If those three clauses are not in your contracts, fix that now.

The companies that win in the synthetic character space will not build the most convincing character. They will build the infrastructure that everyone else needs to build any character at all.

George Lucas did not get rich from Star Wars acting. He kept the merchandising rights and built ILM. The character was the proof of concept. The studio was the business.

Andy Serkis did not build his empire by playing Gollum. He built it by owning the facility and the technique.

Proxy Studios should not be in the business of Tilly Norwood. The intelligent play is building the infrastructure layer that migrates into the broader industry regardless of whether Tilly succeeds as a specific character.

Three business models exist for AI character studios.

Any operator, investor, or brand partner evaluating an AI character studio needs to know which model they are actually funding. The pitch deck will almost always describe all three simultaneously.

Every cycle of synthetic character technology produces the same two outcomes for the same two reasons.

Companies that confused the character with the technology lost. Square closed. Aki disappeared. The technology migrated to competitors who never took the original risk.

Companies that treated the character as a proof of concept for the infrastructure won. ILM. Weta. Imaginarium. The pipeline, not the performer. The technique, not the talent.

The AI character debate is not a debate about whether synthetic performers will be part of the industry. They already are. The debate is a negotiation over who captures the value when the technology matures. That negotiation is happening on social platforms where audience habituation is already underway, in capital markets where the infrastructure bets are being placed, and in institutional frameworks where the boundaries are still being drawn.

The operators who understand all three arenas are building businesses that will still be standing when today's backlash becomes tomorrow's nostalgia.

The character may not survive. The infrastructure will. Build accordingly.

Subscribe to The Creator Operator By Rodrigo Abdalla.

The Operating Manual for Seven-Figure Creator Businesses.