AI companies already consumed the entire internet, and now they need YOU

18 Aug 2026 · 1 min

Watch on YouTube

0:36

In this segment, Dan Novaes argues that major AI firms have effectively "sucked up" the entire internet to train their models, which explains both the impressive capabilities and the current shortcomings of AI systems. He highlights that despite the massive data ingestion, models still lack historical depth and exhibit various dumb behaviors. Novaes then points to a wave of legal actions, noting that Anthropic settled lawsuits with the Wall Street Journal and that Reddit is now licensing its data to AI developers. He emphasizes that the data powering companies like Anthropic and OpenAI originates from everyday consumers, yet these users receive no compensation or benefit from the commercial exploitation of their generated content.

Chapters

The market is completely shifting because these AI companies essentially have sucked up the entire internet. They sucked every single thing up. And that's why the models are pretty smart today, but they're still kind of dumb in many ways, and it's improving day by day, but it doesn't have more data from past times to train on. And then on top of that, there are insane lawsuits happening that Anthropic had to settle with companies like the Wall Street Journal and companies like Reddit are now licensing its data. But who's making that data? The consumers are making that data. So consumers are effectively powering the Anthropics and the OpenAIs of the world. And and they've gotten nothing, absolutely nothing.