Midcentury passes 2M hours of egocentric human data, banks $15M seed for robot training
CyberRobooo · x · 2026-09-27
NYC physical AI startup Midcentury claims its first-person human behavioral dataset has passed 2 million hours, covering 50+ environments and 20,000+ tasks, with hands visible in 90%+ of footage. It includes 3D hand tracking, depth maps, task annotations, plus 50K hours of gameplay and 69K hours of multilingual dialogue. A cloud simulation platform, Matrix, lets humanoids train directly on this data. Citing Dyna Robotics' results that enough egocentric data lets AI grasp physical-world regularities, the milestone helped Midcentury raise a $15M seed — human behavioral data is becoming the "new oil" of physical AI.
Related event: Midcentury Hits 2 Million Hours of Egocentric Human Data for Robotics(2 posts)→
More from Venture
- Solo founder automates ads, content, newsletters, sites and books with AI — TheMoonMidas · 2026-09-27
- AI consultant: listening and requirements skills matter more than the tech — alexcovo_eth · 2026-09-27
- Account grower says AI agents now run his dormant 100k-follower accounts endlessly — TheMoonMidas · 2026-09-27
- Goldman: Hyperscalers would still need $920B/year for depreciation in worst-case capex scenario — SumitGup · 2026-09-27
- Lightspeed partner after a year of AI assistant research: families, not individuals, are the ideal customer — FinanceYF5 · 2026-09-27
- Snorkel AI raises $350M at $3.5B valuation as ARR jumps 18x to $375M — 创业邦 · 2026-09-27