🤖 A.I. - Some Jobs Are Gone. Others Are About to Explode.
- Gato Scatena
- Apr 2
- 8 min read
The New AI Workflow in Filmmaking (And Where You Fit In)
There’s a version of this conversation where people pretend AI is just another tool. It’s not.
What we’re watching right now is a fundamental restructuring of how films get made—from the first idea all the way through international distribution. And like every major technological shift this industry has seen, there are going to be clear winners, clear losers, and a very large group of people who realize too late which side they’re on.
With the introduction of sound into motion pictures, there were plenty of naysayers hesitant to adopt the new technology. The result: Fox overextended and amidst a ton of debt got swallowed by 20th Century; Laemmle lost control of Universal; First National was acquired by WB… and those were just the big boys. Plenty of smaller companies suffered and died.
So the mistake you hardline traditionalists may be making is thinking this is a future problem. It’s not. It’s already here, and it’s already changing how real deals are getting done.
The First Domino: AI Is Already in the Development Room
Let’s start with where most people are quietly using it—development.
Tools like ChatGPT and Claude are already being used across the industry for ideation, and in some cases, a lot more than that. We already know streamers like Netflix have openly acknowledged internal approval for early-stage creative use.
And here’s the part people don’t want to say out loud: writers are absolutely using this to draft scenes. Not final drafts—not yet—but strong enough first passes that the role of the writer, in many cases, is shifting from creator to editor, at least in early workflows.
Internally, we’ve already been using AI in a way that would have sounded aggressive even a year ago. When scripts come in, we’re running them through AI systems that analyze pacing, estimate below-the-line budgets, flag key variables like limited-locations or heavy SFX, and identify characters that might be candidates for upcasting.
We are still reading every script ourselves that’s in contention for financing or production, but if (i) a script doesn’t grab me in the first 15 pages (most would say 5 to 10), and (ii) the AI throws red flags, it might be cooked unless I’m bored. After all, we get a ton of submissions… so I can only imagine how much time and money this process would—read: is—saving companies like Neon and A24.
As for the contenders, AI is giving us and others a head start. We’re walking into scripts already thinking about structure, cost, and casting strategy before we even turn page one.
And so far, it’s been uncomfortably accurate.
The Real Disruption Isn’t Writing—It’s Everything After
Most of the conversation around AI fixates on writers. If anything, it’s my belief that AI is going to enhance professional writers’ output and speed up drafts, proofreading, and ideation. For those keeping it old school, I think they’ll struggle to get footing versus writers now producing four times the output at comparable quality.
The real impact is downstream.
In post-production, tools from companies like Luma AI are already approaching a level where high-end visual effects can be generated at a fraction of the traditional cost. Not perfect yet, but already good enough to break the economics.
Editing is next. We’re not far off from AI-assisted first cuts of feature films that meaningfully reduce both time and cost.
Then you get into distribution—and this is where it becomes unavoidable.
Dubbing is being transformed in real time. AI can now modify jaw and lip movements to match localized dialogue (Flawless AI, et al). Subtitling, which used to cost hundreds or thousands per territory, is quickly becoming a near-zero-cost line item.
And once voice replication becomes normalized, you’re looking at a world where even legacy dubbing voices—or their estates—can continue generating revenue indefinitely without stepping into a recording booth.
This isn’t theoretical. It’s already being used—and most people don’t even realize it.
The Celebrity Question No One Has Answered Yet
There’s another layer here that feels further out—but it’s not as far as people think: AI actors.
Not just digital doubles for stunts or de-aging, but fully synthetic performances tied to real actors’ likenesses and voices. On paper, it sounds like a goldmine. A recognizable face could appear in multiple projects simultaneously without ever stepping on set.
But there’s a problem: scarcity is what creates value. The moment an actor becomes too accessible—too present—their value drops. Audiences get fatigued. The brand weakens.
So while the technology is coming, there’s a natural limiter built into the economics of celebrity itself. This won’t be a takeover—but it will absolutely become some part of the business model after a few companies break the cherry.
🔒 PAYWALL — Where This Actually Gets Dangerous (and Profitable)
Most people are asking, “How do I use AI?”
That’s the wrong question.
The real questions are: where is this creating leverage, where is it quietly destroying value, and where is money about to shift next?
Below is what we’re seeing right now—not theory:
💰 Where margins are expanding (and who’s actually keeping the upside)
⚖️ The legal exposure that could unwind entire libraries
🎬 The animation gold rush—and why it’s about to get crowded fast
🧠 The rise of the “prompt artist” and the compression of entire departments
🌍 The new AI-driven international arbitrage model
If you’re producing, financing, acting, directing, crewing, or selling films, this is where decisions stop being theoretical and start impacting your bottom line. So, let's get into it...