There’s a lot of noise right now around “fully AI-generated films.” I think that misses the more interesting , and much closer reality.
The most viable near-term model isn’t no actors.
It’s one or two central human actors, with AI handling everything else.
Here’s why this hybrid approach works:
• A single real performance anchors emotional truth.
Audiences subconsciously lock onto breath, timing, micro-expression. Once that anchor exists, AI-generated environments and supporting characters become far more acceptable.
• It collapses the hardest AI problems.
Multi-actor blocking, group dialogue timing, long-range emotional coherence, most of these vanish when the human performance is the spine of the film.
• Scope decouples from budget.
AI can generate locations, weather, extras, camera moves, and post-production without a traditional VFX pipeline. You shoot the actor once, then build the world around them.
• It preserves acting as a premium skill.
Rather than replacing actors, it concentrates value. Fewer roles, but more weight. Performance becomes the watermark of authenticity.
This isn’t theoretical. With a locked script, this model is already technically achievable today, and commercially viable within the next 12–24 months, especially for contained thrillers, sci-fi, horror, and character-driven dramas.
I suspect the first mainstream “AI-native” features won’t announce themselves as AI films at all. They’ll just look like ambitious indie features with suspiciously large worlds and suspiciously small crews.
I'm curious how others here see this playing out, especially writers, directors, and actors.
Is this a threat, a tool, or the next normal?
2 people like this
You can use all the AI that you want, if you don't have an interesting, must I say daring, concept it's useless.
The question is, why use A.I.? The "few actor" approach is already possible with old school VFX. For example The Ningyo (https://www.youtube.com/watch?v=T-IFrLvALUc) has extremely high production value, but mostly filmed by a couple in their own house. The Tragedy of MacBeth is using Hollywood actors, but filmed almost entirely at sound stages, with the VFX done by a small boutique. Unless A.I. can achieve something similar or better, in a shorter time, and with a smaller budget, what is the advantage?
We, Writers, Directors, Actors, Composers, and Editors are the creatives with premium skills. These premium skills will be the most difficult to automate.
In the future(maybe near, maybe far), an Auteur Filmmaker who has worn all of these creative hats to make thier "vision" a film, will tap into an AI filmmaker pipeline, and work through all the new "AI problems" ... and make something irresistible just for thier intended audience, and voila. We are in a new era.
I personally thrive on any technological improvement to filmmaking because it democratizes and economizes my "vision" as the storyteller. Nearly everyone who has ever worked in Hollywood has a script that went nowhere. They could not find a champion for thier script, so they went to work in Hollywood ... but soon, everyone with an unchampioned script can become their own champion of thier own "vision" and make thier AI film.
Will premium quality Hollywood filmmaking be abandoned to accommodate streaming bandwidth? Will Hollywood sell more movie tickets and keep thier premium industry alive? Will those who have worked in Hollywood honor the prestige of Hollywood filmmaking when they make thier own AI film? They are the only ones left who might... either way, we are all entering a new era of storytelling!
As an agent and manager, aren't you cutting off work for your clients?
Patrik Gyltefors you ask a valid question Why? It has the same drivers as music production. Cheaper, quicker with still appreciable quality.
I'm not suggesting it is a good thing. We would end up with a plethora of home-generated production, like we have with music. Every second music artist has their own mini music studio in the corner of their lounge room.
For AI generated film, it needs to be cheaper, quicker with just as good viewer quality. Maybe we end up with every second Director, Actor or Producer with a mini studio in the corner of their lounge room.
Home music production has killed off 'mega bands', music studios and many music labels. The creativity is still there, but the money is now in live concerts.
How does that lived experience translate to film?
Philip Sedgwick I see it more like preparation for an inevitable future. As a book publisher, our client base is not Actors, Directors or Producers. Could we migrate into film production? It is worth a look, but I fear there will be no money in it, just like AI music production.
Trevor Learey It's already playing out, but it's far more sophisticated than the retail level generative AI market suspects. As for performance, AI "human" performances are destined to alienate an audience regardless of the beauty of the picture. The "Uncanny Valley" effect which was observed in robotics in the 1970s has been clinically proven to be experienced by audiences watching AI video - revulsion at the "human" actors which are not human is actually a physical and DNA-programmed. 3D animation circumvents this by mocap and actual eyelines which focus naturally in a scene (which gen AI cannot do well or at all) and very careful animation. Here's an interesting infographic on it: http://aerithai.com/papers/uncanny-valley-in-ai.html
I am not threatened in any way by AI and use it almost every day to organize my rambling first thoughts. Having said that, I find most everything about AI to be bland, lacking understanding of the human condition, and devoid of a sense of humor.
But more to the point of the OP, human background players (in my experience) can add life to those background roles in a way that in the examples I’ve seen, AI does not. AI has a learned, not lived perspective. The difference can be subtle yet incredibly impactful.
In AI film generation, money flows to whoever controls scarcity. AI eliminates the scarcity of production, so value collapses as a result.
So who will make money from this disruption?
1. Compute & Model Owners (the apex)2. IP Owners (franchises, characters, worlds)
3. Distribution & Attention Controllers (Attention becomes the real currency)
4. Legal, Compliance & Trust Gatekeepers
5. Star Humans (but fewer of them)
6. Tool operators like storyboarding and AI development will be paid wages.
As a result, our focus is on IP ownership. I don't like it, but I'm not going to ignore.
Trevor Learey, I disagree that A.I. delivers "cheaper, quicker with still appreciable quality." Sure, top notch VFX is expensive,. but that is because of the amount of iteration that goes into it to deliver a highly polished result. Quick mockups, that are still better quality than what A.I. produces, can be at almost zero budget (as an example, I did this video in 10min, while live streaming the process on my YouTube channel: https://www.stage32.com/media/3977699606556321845).
1 person likes this
Patrik Gyltefors What I meant is that AI will need to be cheaper, quicker with some quality. I'm not suggesting it is there yet, or even close.
1 person likes this
Paul Rivers, we are already there. Never before has it been so easy for anyone to create their own film. Everyone has a decent (while not "perfect") camera in their pocket (iPhones etc.), free access to motion capture libraries like Mixamo, an infinite amount of photo-scanned assets from Quixel Megascan, realtime cinematic rendering with Unreal...
It is inevitable that AI will continue to integrate into art. It is up to the filmmaker to work with it responsibly.
Amanda Toney The challenge over the next decade at least, is to use AI where it saves time and money, but maintain old school techniques for authenticity and integrity.
The audience has a very attuned BS filter, so AI integration should be very gradual as quality improves. I hear it with AI music. Almost too good to have sole. Lacks human unpredictability and mystery.