No Need For Tilly to Have Us in a Tizzy – “She” is Just a New Tool For Storytelling
by James Mathers
Cinematographer and Founder of the Digital Cinema Society
(Excerpted from the October 2025 Digital Cinema Society eNewsletter)
By now I’m sure you’ve heard about “AI-Actress” Tilly Norwood via a brilliant publicity stunt by the virtual character’s creator, AI studio Particle6’s CEO Eline Van der Velden. She announced that she was seeking representation for Tilly as an actress from top entertainment industry talent agents. The media frenzy was predictable with a chorus of well known actors, including Natasha Lyonne, Toni Colette, and Whoopi Goldberg, up in arms. Emily Blunt exclaimed in an interview with Variety, “Good Lord, we’re screwed.”
Meanwhile, SAG-AFTRA released a statement that, “Tilly Norwood is not an actor, it’s a character generated by a computer program that was trained on the work of countless professional performers — without permission or compensation.” They went on to say, “it creates the problem of using stolen performances to put actors out of work, jeopardizing performer livelihoods and devaluing human artistry.” They warned, “signatory producers should be aware that they may not use synthetic performers without complying with our contractual obligations, which require notice and bargaining whenever a synthetic performer is going to be used.”
Tilly’s creator, Eline Van der Velden, countered with the statement, “[Tilly] is not a replacement for a human being, but a creative work – a piece of art. Like many forms of art before her, she sparks conversation, and that in itself shows the power of creativity.” One thing is for sure, at starting a conversation, Tilly is very effective.
But what is Tilly Norwood? “She” is obviously not an actress, but rather a collection of data points that claims to be an original character because the software is trained on “all available” data as opposed to being modeled as a depiction of one particular person. Since Tilly is a melding of so many disparate depictions of various humans coming from the vast trove of abundantly available imagery on the public web and elsewhere, there is some legal cover. Lawyers would have a field day if Tilly looked or sounded too much like a particular celebrity, say like Scarlett Johansson. The legal cover comes from the amalgamation of likenesses.
There is no doubt that this software creation, or amalgam of data creations, should be licensable, just like you might license other software in the App store, but just don’t put it on the same pedestal as a human performance. After all, real acting is re-acting; an authentic performance can only come from responding truthfully to other actors and the environment, not from pre-planned or forced expressions. I don’t think Tilly or other similar algorithmic depictions of humans will be much use as a lead character in a narrative production. However, that is not to say that such characters will not find use in mass media. A product spokesperson delivering a pitch to camera could work, and that will certainly rob real actors of some livelihood, but if the audience is not discerning enough to notice the difference, then is there really a difference? And I’m sorry to say so to my stunt performer friends, but it does make a lot of sense for an AI likeness of an actor to be put in harm’s way instead of a lead or stunt double.
We have been combining humans and computer generated illusions for decades. For the 2019 movie, Gemini Man, for example, director Ang Lee used a hyper-realistic CGI version of a Will Smith double, (albeit 30 years younger,) to create one of the best car/motorcycle chase scenes I’ve ever seen; one that would have been insanely dangerous, or completely impossible by traditional stunt methods. 
In this example, and dozens of others that I could name, the CGI performance was driven by mo-cap, so live actors were creating the performance. AI is now adding the ability for those illusions to be driven simply by text prompts drawing from the combined past performances of countless actors.
Tilly Norwood is only one evolutionary step to where this kind of imaging technology is headed. Some may choose to fight it while others will attempt to find new ways to use this technology for storytelling. If you want to keep up, a good example of the state of this technology, including scenes with Tilly, can be found in their demo reel on the Particle6 website:
https://www.particle6.com/
I enjoyed your article. I couldn’t agree more-this new AI actress, Tilly Norwood, is just another step forward in the long line of tech shifts we’ve seen in this business.
It actually took me back to a screening I attended in Hollywood around 2000 or 2001, where several cinematographers showed side-by-side tests comparing 35mm film with the new digital cameras of the time-the Panasonic VariCam and the Sony F900, if I remember right.
I was fresh out of film school then, sitting in a huge theater full of DPs, including John Toll, who I had the opportunity to chat with prior to the screening. I remember feeling the tension in the packed theater-all these working cinematographers were a little worried, and I was too. I’d spent the last few years learning film, and they’d all spent their careers mastering it. Suddenly, it felt like we might have to start over.
But of course, we didn’t start over. We adapted. We learned the new tools and kept shooting, just like we always do. What you said in your article really hits home-it’s all just technology. It changes, we change with it, and the storytelling goes on. The art never disappears; only the tools evolve