Netflix has just handed over $587 million for InterPositive, an AI-native filmmaking startup co-founded by Ben Affleck. TechCrunch has the details, and I'd encourage anyone who builds or markets content for a living to read it properly, not just the headline.
Because the headline is easy to misread as "Hollywood does AI slop now." That's not what's happening. What's happening is a lot more interesting, and a lot more useful to those of us running smaller operations.
What InterPositive actually is
Strip away the celebrity name and what Netflix bought is a pipeline — pre-visualisation, shot planning, VFX assist, dialogue and ADR cleanup, and a rendering layer that cuts weeks off post-production. It's not a tool that generates a film from a prompt. It's automation stitched into every stage where human time used to be the bottleneck.
That distinction matters enormously. I've spent the best part of four decades writing software, and the pattern is always the same: the first wave of hype is "AI replaces the creative," and the money that actually gets made is in "AI removes the friction around the creative." Netflix didn't buy a director. They bought a workflow.
Why a streamer pays nine figures for a workflow
Netflix ships an enormous volume of content. Every day shaved off a production cycle, every render pass that doesn't need a human to babysit it, compounds across hundreds of projects a year. At that scale, $587 million for a pipeline that reliably takes cost and time out of production isn't a punt — it's infrastructure spend, the same category as building your own data centre or owning your own CDN.
That's the validation worth noticing. This isn't a studio dabbling in a gimmick. It's a business case, run through finance, that says owning the automation layer is worth more than renting it from a dozen point solutions. I've been saying versions of that sentence to clients for years, usually about content marketing rather than feature films, and it's satisfying to see it proven at this scale.
Does this trickle down?
Here's the honest answer: the specific tooling won't, not for a good while. InterPositive is built for VFX-heavy, high-budget production and it'll stay locked inside Netflix's stack. You're not getting access to it, and neither am I.
But the underlying idea absolutely trickles down, and it already has. What Netflix just did for feature films at nine-figure scale is exactly what's been happening for two years in content marketing, YouTube, and publishing — just several zeroes smaller. Take raw material — a podcast transcript, an article, a stack of RSS feeds, a book manuscript — and run it through a pipeline that handles the repetitive stages so a human only touches the parts that need judgement.
That's the whole premise behind the tools I've built. Article2Video takes written content and turns it into video without anyone spending a day in an edit suite. RSSMasher takes a flood of feed content and turns it into something publishable at volume. VidMasher and AIMasher exist for the same reason InterPositive exists inside Netflix: because doing it by hand doesn't scale, and doing it badly with a single AI tool bolted on doesn't scale either. You need the pipeline, not the gadget.
The bit people miss
The lesson from InterPositive isn't "AI can make films now." Plenty of tools could already do bits of that. The lesson is that the money follows integration, not novelty. Anyone can wire up an API call to a model. The value sits in the boring middle — the shot planning, the handoffs, the quality control — where a proper pipeline turns a clever demo into something you can run a business on.
If you're building or marketing content and you're still treating AI as a single tool you dip into occasionally, this is your nudge to think in pipelines instead. Where's your raw material coming in? Where does it need a human eye? Where can the repetitive 80% run itself? That's the question InterPositive answered for feature films at a scale most of us will never touch. It's the same question worth answering for your own content operation, at whatever scale you're actually working at.
Raw material in, something worth publishing out. That's the whole game, whether you're spending $587 million or a fraction of that on the right SaaS stack.
— Wayne