The Brainrot Economy

Hi Friends,

For years, the venture world has been obsessed with network effects, proprietary datasets, and high switching costs. Today, as I watch my kids navigate their digital worlds, I’ve started to realize the last defensible barrier in a world of commoditized software might actually be the trust of a five-year-old.

Growing up, shows like Sesame Street, Mr. Rogers, and Barney were the formative pillars of my childhood. They provided a shared language and a sense of discovery. Now, I watch my daughter burn through the same ten YesNeo or Bebefinn episodes on repeat because the supply of content that actually captures her heart is remarkably scarce. In silence, she either retreats to what she’s already seen or drifts toward the YouTube algorithm, a system I keep away from her specifically because it is designed to optimize for the next click.

My son is ten, a leap-day baby who learned division early just to explain his own birthday. He grew up on Numberblocks, a show that taught him to count to 100 before he could tie his shoes. He’s since moved on to advanced math through tools like Synthesis, yet in his downtime, he still wants to watch something. The problem is that Numberblocks has nothing left for him. The characters aged out of his life because the legacy economics of animation didn't allow the content to evolve alongside his curiosity. That is a failure of imagination, dictated by an outdated production model.

At Flybridge, our largest bet is Native AI for Human Potential. We believe AI creates entirely new ways for people to learn, create, and connect. While much of the industry remains focused on enterprise productivity, my kids keep pulling my attention towards media.

Historically, editorially-driven, AI-native media was not possible to build. The tools required to bridge high-end production with real-time relevance simply didn’t exist. That changed this year. 

There is now an opportunity for a completely new category of companies that the venture world hasn't fully priced in yet. I call it vertically integrated, editorially driven, AI-native media.

The Editorial Moat

These companies own their voice the way The Economist or early HBO owned theirs. They use AI to handle production at a cost and speed that legacy media cannot touch, but the technology is just the engine. The actual moat is editorial judgment, accumulated trust with a specific audience, and the feedback loop that sharpens with every time a child watches, skips, or shares.

The incumbents are currently retreating from this challenge. Netflix’s kids' commissions dropped 42% in early 2024. Warner Bros. Discovery gave up Sesame Street after decades of stewardship. These companies are choosing to acquire existing IP rather than create new IP themselves.

Meanwhile, the demand signal is deafening. Ms. Rachel ranked in Netflix's global top 10 for 17 straight weeks. Bluey racked up 842 million hours of US viewing in 2024. Families with children average 13.6 streaming services. Parents are starving for content worth watching, but the market is split into two broken buckets. You have YouTube Kids, an infinite stream of algorithmically optimized junk, and you have Netflix or Disney+, which offer beautiful, developmentally sound shows on an 18-month production cycle.

That cycle makes legacy players structurally unable to respond to the world. When a major event happens, such as a war, an election, or a global shift, there is no Sesame Street segment for a ten-year-old that drops within 48 hours. But AI removes that ceiling.

Scaling Taste

We have reached an inflection point in AI video. Production costs are plummeting, enabling personalization at a scale previously considered impossible. Imagine a show where the references, the food, and the daily problems match the life of a child in Lagos as closely as they do a child in Ohio. Localization can now happen at the cohort level, flowing through parent accounts to ensure privacy and trust.

The IP strategy here is a sequenced bet. You license characters with proven followings to solve the "cold-start" problem, building distribution and a subscriber base. Then, you develop AI-native originals in parallel. The right licensed IP buys the time and the audience to introduce something entirely new.

The hardest thing to build in this new ecosystem will be editorial taste. Ms. Rachel reached 16 million subscribers by understanding the specific mechanics of how toddlers learn. Joe Brumm made Bluey a global phenomenon by respecting both the child's intelligence and the parent's exhaustion. That is taste. It cannot be prompted. It has to be hired.

The people who possess this, like the editors, writers, and child development experts, are currently trapped in organizations that are cutting budgets. Now is the moment to recruit them. 

LLMs tend to flatten a voice into a probabilistic average. The solution is a structure in which humans greenlight model drafts. The LLM is a production tool, while the human remains the publishing authority.

The Opportunity

This logic extends far beyond one use case like childhood media. It applies to sports fans who want to understand the "why" of a specific play, or first-time investors trying to decode a Fed announcement in plain English. In every vertical, the existing content is either too shallow, too slow, or too expensive to keep current.

I’ve spent the last few months meeting founders who are building at this intersection. I want to meet more.

If you’re building an AI-native content platform with real editorial standards, whether in children’s media, sports, finance, or civic media, I want to talk.

DM me or reach me at [email protected].

— Jesse