Nye's Digital Lab is a weekly scribble on technology, culture, and what it means to build things in a world that's changing faster than our contracts.
This week I'm looking at Ben Affleck, a $600 million AI deal, and what it means for every creative who's been generating data for free.
In a morbid landscape, I've been hunting for new business models. How will creatives make money in an automated world?
This week, I believe there is a ray of light.
Netflix just acquired movie star, Ben Affleck's, quietly-built AI company, InterPositive, for a reported sum of up to $600 million.
InterPositive doesn't generate synthetic actors or hallucinate dialogue. It isn't Runway or Luma or some new model from Gemini. It does something smarter and more subversive: it lets filmmakers build their own proprietary AI models from footage they actually shot.
"You can use your own model to remove the wires on stunts, reframe a shot, get a shot you missed, shape the lighting, enhance the backgrounds."
Ben Affleck's motivation is clear:
"The kind of judgment that takes decades to build, experience to hone, and only people can have."
The data from InterPositive is trained on your production, not scraped from the internet, not borrowed from some foundational model trained on everyone else's work. Yours. And that threatens every SaaS AI video platform and every film studio at the same time.
This is your ray of light, creatives. This is the move.
It's not just for Hollywood A-listers with soundstages and Oscar statuettes. The model scales.
Here's how it breaks down.
What made InterPositive valuable wasn't just the tooling. It was the data discipline.
Affleck and a small team filmed a proprietary dataset on a controlled soundstage, deliberately building "smaller datasets and models focused on filmmaking techniques, rather than performances." They were intentional about what they captured and, crucially, who owned it.
This is the creative business model hiding in plain sight: the act of making something professionally is also the act of generating data. Every lighting setup, every camera angle, every retake where the director says "one more time, slightly slower" is digital information. Pattern. Style. Visual logic. Most creatives are generating that data and handing it to someone else for free.
The InterPositive approach flips this. You build a clean, controlled, high-quality dataset from your own production process. You own the model it trains. You license the outputs, or you sell the company to Netflix. A group of actors and writers could do this. A dance company. A stunt collective. A studio of illustrators.
GO. Start today.
The creative act, defended and documented at a professional level, becomes not just art but infrastructure.
The act of making something professionally is also the act of generating data. Most creatives are generating that data and handing it to someone else for free.
The key phrase from Netflix's announcement is worth memorizing: "Innovation should empower storytellers, not replace them." That's not just PR. That's a market signal. Netflix's co-CEO Ted Sarandos said in 2024 that there's "a better business and a bigger business in making content 10% better" than in making it "50% cheaper."
Quality-controlled human creativity, captured as proprietary data, is now a product category. You want to be in that category, not feeding it.
Modern aircraft can take off, navigate, manage turbulence, and land with minimal human input. The autopilot doesn't get tired. It doesn't need coffee. And yet, every airline in the world keeps the pilots. Why?
Passengers don't just want a safe flight. They want to know a human made a judgment call if something went sideways.
The entertainment version of this is what Affleck's play is framing. Audiences are developing a taste for knowing whether a human was at the controls. AI-generated content, the stuff people are calling "slop" online, is becoming distinguishable, and not in a flattering way. There's a growing cultural appetite for the certified-human label, the creative equivalent of a pilot announcement over the intercom.
Seth MacFarlane recently made himself into Bill Clinton using AI tools. Not really my thing, but it was genuinely creative. It revealed something important: people can engage with AI-assisted work when there's a human sensibility steering it, when the joke lands because a person with taste chose to make it. The technology is the instrument. The musician still matters.
And then there's the Cat Kung Fu phenomenon. I watched an embarrassing amount of it this week. It's absurd, algorithmically targeted content, and it works, in the same way a bag of chips works. There is clearly a market for it. But will it sustain the ecosystem it's operating within? Will it pay writers, fund pilots, support studios, maintain the production infrastructure that the next InterPositive will need?
That's the real question.
Boycotting AI slop is a sliver away from demanding the pilot be on board the plane. People want to know a human made the call.
It is my hope that a creative community that maintains high-quality, human-centered production standards and documents them creates something audiences can orient around. The authenticity becomes a feature.
And features, in this market, can be licensed. This is big.
The Affleck playbook isn't about being famous enough to sell a company to Netflix, though it helps to get the jump. It's about a sequence of decisions any serious creative group can make right now.
For every creative out there: start saving your data locally.
Produce at a professional level, deliberately.
InterPositive didn't scrape the internet. They built a soundstage workflow with the "consistency and controls" cinematographers and directors expect. The quality of the input data is the whole game. If you're an actor, a writer's room, a motion capture team, a choreography collective, the equivalent is: do your best work, in a controlled environment, with intentional documentation.
Treat the data as an asset from day one.
Think about consent, ownership, and licensing before the shoot, not after. InterPositive built in "restraints to protect creative intent," keeping creative decisions with the artists and ensuring "the benefits of this technology flow directly back to the story they're trying to tell."
That kind of structural thinking, data governance baked into the creative process, is the difference between a dataset you own and one you accidentally donated.
The question isn't only "what are we making?" It's "what are we building, and who owns the infrastructure?"
Will this sustain the world it's creating? Likely not entirely, but it's an indication of the direction we need to sprint toward. A creative dataset trained on professional human work, owned by the humans who made it, is more durable than any content deal. It compounds. It can be licensed, refined, and defended legally in ways a streaming output deal cannot.
I find this quote from Affleck interesting:
"From the invention of the moving image to the transition to digital, from motion capture to virtual production, technology has evolved alongside the artists who use it."
Every one of those transitions had winners and losers. The winners were usually the ones who understood early that the new tool was also a new kind of property.
People still want the pilot. The culture still wants to know a human made the call. That instinct is your market. The dataset is your moat. And the sprint starts now.
Seriously, go!
Make it happen.
Hey! That’s it for this time. I do this every week, if you vibe to the ideas I express, consider subscribing or sharing with friends. If you like tech-detoxing with a book like I do, I crammed some of last years best essays into a printed collection.
This essay was built from my personal knowledge base, and a messy collection of agents crawling over my Obsidian / Claude Code infrastructure. Final sweetening was done with Claude Sonnet 4.5. Images created in Flow State with Leonardo.ai.
For more info visit: https://nyewarburton.com
We’ll see you next time.
Netflix acquisition announcement and Affleck's statement on InterPositive's mission: about.netflix.com
Deal valuation reported at up to $600 million (Bloomberg, via Variety): variety.com
Full acquisition breakdown and background on InterPositive's stealth development: deadline.com
Ted Sarandos "10% better" quote (2024): deadline.com
NPR coverage including Creators Coalition on AI context: npr.org
Hollywood Reporter on InterPositive's origins and the Artists Equity first-look deal: hollywoodreporter.com
Nye's Digital Lab is a weekly scribble on technology, culture, and what it means to build things in a world that's changing faster than our contracts.
This week I'm looking at Ben Affleck, a $600 million AI deal, and what it means for every creative who's been generating data for free.
In a morbid landscape, I've been hunting for new business models. How will creatives make money in an automated world?
This week, I believe there is a ray of light.
Netflix just acquired movie star, Ben Affleck's, quietly-built AI company, InterPositive, for a reported sum of up to $600 million.
InterPositive doesn't generate synthetic actors or hallucinate dialogue. It isn't Runway or Luma or some new model from Gemini. It does something smarter and more subversive: it lets filmmakers build their own proprietary AI models from footage they actually shot.
"You can use your own model to remove the wires on stunts, reframe a shot, get a shot you missed, shape the lighting, enhance the backgrounds."
Ben Affleck's motivation is clear:
"The kind of judgment that takes decades to build, experience to hone, and only people can have."
The data from InterPositive is trained on your production, not scraped from the internet, not borrowed from some foundational model trained on everyone else's work. Yours. And that threatens every SaaS AI video platform and every film studio at the same time.
This is your ray of light, creatives. This is the move.
It's not just for Hollywood A-listers with soundstages and Oscar statuettes. The model scales.
Here's how it breaks down.
What made InterPositive valuable wasn't just the tooling. It was the data discipline.
Affleck and a small team filmed a proprietary dataset on a controlled soundstage, deliberately building "smaller datasets and models focused on filmmaking techniques, rather than performances." They were intentional about what they captured and, crucially, who owned it.
This is the creative business model hiding in plain sight: the act of making something professionally is also the act of generating data. Every lighting setup, every camera angle, every retake where the director says "one more time, slightly slower" is digital information. Pattern. Style. Visual logic. Most creatives are generating that data and handing it to someone else for free.
The InterPositive approach flips this. You build a clean, controlled, high-quality dataset from your own production process. You own the model it trains. You license the outputs, or you sell the company to Netflix. A group of actors and writers could do this. A dance company. A stunt collective. A studio of illustrators.
GO. Start today.
The creative act, defended and documented at a professional level, becomes not just art but infrastructure.
The act of making something professionally is also the act of generating data. Most creatives are generating that data and handing it to someone else for free.
The key phrase from Netflix's announcement is worth memorizing: "Innovation should empower storytellers, not replace them." That's not just PR. That's a market signal. Netflix's co-CEO Ted Sarandos said in 2024 that there's "a better business and a bigger business in making content 10% better" than in making it "50% cheaper."
Quality-controlled human creativity, captured as proprietary data, is now a product category. You want to be in that category, not feeding it.
Modern aircraft can take off, navigate, manage turbulence, and land with minimal human input. The autopilot doesn't get tired. It doesn't need coffee. And yet, every airline in the world keeps the pilots. Why?
Passengers don't just want a safe flight. They want to know a human made a judgment call if something went sideways.
The entertainment version of this is what Affleck's play is framing. Audiences are developing a taste for knowing whether a human was at the controls. AI-generated content, the stuff people are calling "slop" online, is becoming distinguishable, and not in a flattering way. There's a growing cultural appetite for the certified-human label, the creative equivalent of a pilot announcement over the intercom.
Seth MacFarlane recently made himself into Bill Clinton using AI tools. Not really my thing, but it was genuinely creative. It revealed something important: people can engage with AI-assisted work when there's a human sensibility steering it, when the joke lands because a person with taste chose to make it. The technology is the instrument. The musician still matters.
And then there's the Cat Kung Fu phenomenon. I watched an embarrassing amount of it this week. It's absurd, algorithmically targeted content, and it works, in the same way a bag of chips works. There is clearly a market for it. But will it sustain the ecosystem it's operating within? Will it pay writers, fund pilots, support studios, maintain the production infrastructure that the next InterPositive will need?
That's the real question.
Boycotting AI slop is a sliver away from demanding the pilot be on board the plane. People want to know a human made the call.
It is my hope that a creative community that maintains high-quality, human-centered production standards and documents them creates something audiences can orient around. The authenticity becomes a feature.
And features, in this market, can be licensed. This is big.
The Affleck playbook isn't about being famous enough to sell a company to Netflix, though it helps to get the jump. It's about a sequence of decisions any serious creative group can make right now.
For every creative out there: start saving your data locally.
Produce at a professional level, deliberately.
InterPositive didn't scrape the internet. They built a soundstage workflow with the "consistency and controls" cinematographers and directors expect. The quality of the input data is the whole game. If you're an actor, a writer's room, a motion capture team, a choreography collective, the equivalent is: do your best work, in a controlled environment, with intentional documentation.
Treat the data as an asset from day one.
Think about consent, ownership, and licensing before the shoot, not after. InterPositive built in "restraints to protect creative intent," keeping creative decisions with the artists and ensuring "the benefits of this technology flow directly back to the story they're trying to tell."
That kind of structural thinking, data governance baked into the creative process, is the difference between a dataset you own and one you accidentally donated.
The question isn't only "what are we making?" It's "what are we building, and who owns the infrastructure?"
Will this sustain the world it's creating? Likely not entirely, but it's an indication of the direction we need to sprint toward. A creative dataset trained on professional human work, owned by the humans who made it, is more durable than any content deal. It compounds. It can be licensed, refined, and defended legally in ways a streaming output deal cannot.
I find this quote from Affleck interesting:
"From the invention of the moving image to the transition to digital, from motion capture to virtual production, technology has evolved alongside the artists who use it."
Every one of those transitions had winners and losers. The winners were usually the ones who understood early that the new tool was also a new kind of property.
People still want the pilot. The culture still wants to know a human made the call. That instinct is your market. The dataset is your moat. And the sprint starts now.
Seriously, go!
Make it happen.
Hey! That’s it for this time. I do this every week, if you vibe to the ideas I express, consider subscribing or sharing with friends. If you like tech-detoxing with a book like I do, I crammed some of last years best essays into a printed collection.
This essay was built from my personal knowledge base, and a messy collection of agents crawling over my Obsidian / Claude Code infrastructure. Final sweetening was done with Claude Sonnet 4.5. Images created in Flow State with Leonardo.ai.
For more info visit: https://nyewarburton.com
We’ll see you next time.
Netflix acquisition announcement and Affleck's statement on InterPositive's mission: about.netflix.com
Deal valuation reported at up to $600 million (Bloomberg, via Variety): variety.com
Full acquisition breakdown and background on InterPositive's stealth development: deadline.com
Ted Sarandos "10% better" quote (2024): deadline.com
NPR coverage including Creators Coalition on AI context: npr.org
Hollywood Reporter on InterPositive's origins and the Artists Equity first-look deal: hollywoodreporter.com
Nye's Digital LabTech's Perfect Storm of Layoffs
Covid's Whiplash, India & Billions of AI Cloud Infrastructure
Nye's Digital LabChildhood's End
The science fiction story that keeps me up at night
Nye's Digital LabLEGO to Protocol: Part I
A LEGO two-parter on the future of modular global optimization
Nye's Digital LabTech's Perfect Storm of Layoffs
Covid's Whiplash, India & Billions of AI Cloud Infrastructure
Nye's Digital LabChildhood's End
The science fiction story that keeps me up at night
Nye's Digital LabLEGO to Protocol: Part I
A LEGO two-parter on the future of modular global optimization
Weekly scribbles on creativity in the age of AI & distributed systems.
Weekly scribbles on creativity in the age of AI & distributed systems.
Share Dialog
Share Dialog
Subscribe to Nye's Digital Lab
Subscribe to Nye's Digital Lab
>200 subscribers
>200 subscribers
No activity yet