1X NEO home robot launches at $20K — teleoperated by US employees, not autonomous AI

Summary

1X Technologies launched NEO, a $20,000 humanoid home robot available for preorder at $499 per month. The robot has generated significant social media debate around a single issue: it is teleoperated, not fully autonomous.

1X's FAQ states that NEO "works autonomously by default" but for unfamiliar tasks, users can "schedule a 1X expert to guide it." These experts are 1X employees based in the USA. The robot learns from teleoperated sessions and improves over time. Wall Street Journal reviewer Joanna Stern confirmed that 100% of NEO's actions were teleoperated in her demo. MKBHD framed this as a flaw, but the hosts treat teleoperation as part of the business model rather than a product failure.

The real tension is economic. 1X manufactures in California and hires remote operators at US wages, far above the $2-per-hour figures some observers cited online. Unit economics are rough initially, meaning the company burns significant capital per robot sold. This mirrors the venture capital pattern at Waymo (founded 2009, still unprofitable 16 years later), Reality Labs (ongoing losses), and early OpenAI (five years of losses before revenue). Frontier robotics requires sustained losses for a decade or more. The question is whether 1X stays capitalized long enough to reach profitability.

Boston Dynamics went enterprise, selling to DHL, BP, and energy companies, and is now 80% owned by Hyundai. Figure AI claims end-to-end autonomous learning for household tasks but hasn't allowed independent reviewers to test it and hasn't opened preorders. 1X is shipping aggressively, already taking $20,000 deposits and positioning itself as moving faster than Figure toward actual delivery.

1X targets "cute"—the NEO has deliberately soft aesthetics, contrasting with Tesla's Optimus, which Elon Musk frames as "cool and badass." The cute design becomes awkward in certain contexts. One host noted the design reads as "fundamentally a little bit demonic" without a mouth, and adding one risks uncanny valley. Another host saw an eight out of ten on cuteness until context shifts it: a humanoid holding a kitchen knife in your home becomes frightening.

Preorders are bets on improvement. Buyers are essentially funding training data collection. Early adopters should treat NEO like the Oculus DK2, a developer kit rather than a finished product. One host compared NEO's early performance to a helpful four-year-old: slow, distracted, prone to dropping things. Expect messy but improving performance over months or years.

There is also a labor arbitrage layer. If teleoperation becomes routine, a Los Angeles resident paying premium rates for house cleaning could theoretically be serviced by an operator elsewhere in America or globally. This becomes a supply shock that drives prices down, and a political issue. Bernie Sanders already criticized Friend.com despite its limited market penetration. Robots are now politically salient, even if no single player dominates. 1X's choice to hire US-based operators is deliberate optics. International teleoperation would invite immediate political backlash.

One speculative model: gig workers could operate NEO units between tasks, similar to Uber drivers picking up rides. A single robot could hand off between multiple operators, with learning happening across a distributed network of handlers. This hasn't been announced but fits the unit economics.

Preorder pricing creates a lease-versus-buy tension. At $499 per month, buying outright ($20,000) only makes sense if NEO stays competitive for 40 months. Given how fast robotics is advancing, leasing may be rational—upgrade every few years rather than lock in capital. 1X hasn't committed to a clear upgrade path.

1X's messaging has been consistent and honest. It hasn't overpromised, differentiating it from Rabbit R1 and Humane AI Pin, which generated hype, shipped incomplete products, and disappointed buyers. MKBHD's skepticism stems partly from that track record. His warning—this is a preorder for an unfinished product, currently 100% teleoperated, know what you're buying—is fair consumer advice, even if others view teleoperation as a feature rather than a flaw.

The broader story is less about NEO itself and more about 1X's execution model: ship something real, take money, iterate with live customer feedback, burn capital for years, and hope to reach scale before fundraising dries up. Whether 1X stays the course depends partly on external capital (rumors suggest a $10 billion fundraise may be coming) and partly on Figure AI's performance. Figure is valued near $40 billion and faces pressure to prove autonomous household robots work. If Figure succeeds, it validates the category. If it stumbles, 1X's methodical teleoperated approach looks more realistic than the competition.