Moonshots EP 188: Humanoid Robots Are Coming to Your Home This Decade

Summary

On-site interview at 1X Technologies’ Palo Alto facility with CEO and founder Bernt Bornich, joined by Dave Blundin. Bornich explains 1X’s strategic bet on the home market over industrial: consumer hardware scales differently (iPhone reached 1.7B units via near-doubling annual growth), and intelligence requires data diversity that factory repetition cannot provide — their previous-gen EVE robots plateaued after 20-40 hours of learning in logistics/guarding tasks. The Neo Gamma is designed from first principles for home deployment: 66 lbs, 5’4”, can lift 150 lbs (athletic human strength-to-weight ratio), 22 degrees of freedom per hand matching human anatomy, completely silent operation, soft/huggable exterior, 4-hour battery with fast recharging. Price target is around $30K purchase or ~$300/month lease ($10/day, $0.40/hour). Bornich frames the company as pursuing AGI through embodied learning rather than just applying labor — arguing that spatial/temporal intelligence is more fundamental than language, and that 10,000 deployed home robots would generate more useful novel data daily than non-duplicate YouTube uploads. Manufacturing roadmap: north of 20K annual run rate by end of 2026, with order-of-magnitude jumps afterward, though aluminum sourcing and assembly labor become constraints at iPhone-scale volumes. Bornich sees the “hard takeoff moment” as robots building robots, building data centers, chip fabs, and energy infrastructure. He builds AI in-house rather than using hyperscaler models, believing embodied world models trained on real interaction data (observation + action + goal) will surpass language-first approaches.

Key Segments

• [00:00-04:00] Tour of 1X facility, live Neo Gamma interaction, Peter secures first home unit
• [04:00-07:00] Why home over factory: consumer scale, data diversity, EVE plateau at 20-40 hours
• [10:00-15:00] Design philosophy: safe, capable, affordable; few hundred parts vs car’s 50K; price target ~$30K
• [17:00-22:00] In-house AI strategy: spatial/temporal intelligence over language-first; embodied learning loop (theory-action-observation) vs passive internet data
• [22:00-27:00] Manufacturing roadmap: 20K/yr by end 2026, aluminum sourcing limits at scale, robots needed to build robots

Notable Claims

• Neo Gamma: 66 lbs, lifts 150 lbs, 22 DOF per hand, 4-hour battery, half-hour recharge
• 10,000 deployed home robots would generate more useful daily data than non-duplicate YouTube uploads
• Previous-gen EVE robots plateaued learning after 20-40 hours on repetitive tasks
• Closer to a “very complicated refrigerator” than a car in manufacturing complexity (hundreds of parts vs 50K)
• iPhone manufacturing displaced large portions of Chinese labor force and still ran out, expanding to neighboring countries
• Factory run rate target: north of 20K annual by end of 2026