Unitree IPO value slumps as robot brains lag

The physical AI sector is at a critical juncture. The recent IPO of Unitree and its subsequent value drop highlight the gap between hardware progress and software intelligence. While the event at Actuate showed signs of growth, the crisis of data quality remains a bottleneck. Developers are exploring various approaches, from task-specific robots that generate real-world data to general models that aim for versatility but face reliability issues. The entry of auto companies like Tesla, Wayve, and Uber into humanoid robotics suggests a convergence of expertise in ML tooling and vehicle data, which could accelerate progress. However, the debate over whether to focus on verticals or go general is unresolved, with leaders like Gervet cautioning that narrow solutions may be crushed by more advanced models. The future 'ChatGPT moment' for physical AI might not come from a single breakthrough, but rather from a gradual improvement in reliability and distribution, as Foxglove's CEO suggests. Ultimately, the sector is waiting for a product that excites consumers and proves value beyond investor enthusiasm.