The AI race everyone’s watching is only half of the actual story…

24 Aug 2026 · 0 min

Watch on YouTube

0:25

In this brief monologue, Dan Novaes critiques the prevailing focus on compute resources—chips, data centers, and energy—as the primary driver of AI scaling. He argues that this perspective captures only half of the equation, emphasizing that without continual human-generated data, AI models cannot improve. Novaes quantifies the contribution of people, stating they constitute roughly one‑third of the factors influencing AI progress and are likely the most critical component. The overall message is that the AI race narrative is incomplete without recognizing the central role of human data and participation.

Compute is only half the battle in scaling AI. Everyone's focused on chips, everyone's focused on data centers, energy, and you're starting to see that in the news cycles. That's really only half of what's important on this. Without human data, nothing's going to happen. It's like we are stuck, right? So it needs to train on new information. And the reason these models get better is because it gets more human data. People are one-third of the equation for this, and it's by far and away probably the most important.