OpenAI-backed robotics startup 1X has unveiled a groundbreaking AI world model that eliminates the need for human operators to train its humanoid robot, Neo. The announcement, made Monday by CEO Bernt Børnich, marks a significant shift in how humanoid robots learn to perform tasks.
Traditionally, humanoid robot companies have relied on armies of human operators wearing VR headsets, motion-capture suits, and sensors to demonstrate tasks ranging from squatting to washing dishes. These teleoperation sessions generate training data that teaches robots how to navigate the physical world. However, 1X’s new approach allows Neo to learn directly from video captured by the robot itself, fundamentally changing the data collection paradigm.
“The big unlock is essentially now that intelligence scales with the number of deployed robots, instead of the number of operators you have gathering data,” Børnich explained. The world model—an AI system capable of simulating realistic environments and real-world physics—performs the same function as human operators would, while improving Neo’s ability to generalize and tackle unfamiliar tasks.
This shift addresses significant concerns about robot training work, which operators have described as physically demanding and tedious. Some Tesla Optimus trainers have reported sustaining injuries, and positions typically pay around $25 per hour at both Tesla and 1X. A 1X spokesperson confirmed the new world model “significantly reduces” reliance on teleoperation, with robots expected to collect most data themselves going forward.
Neo, priced at $20,000 or $500 monthly subscription, captured internet attention last October with a viral 10-minute demo showing it vacuuming, folding laundry, and unloading dishwashers. The robot is expected to ship this year, though early adopters should expect some remote human operation initially. Børnich projects Neo will perform most tasks autonomously by year-end, with fully autonomous operation expected by 2026.
1X, which raised $100 million in 2024 with backing from OpenAI and Samsung, plans to produce over 10,000 robots this year. The company reportedly “sold out in the first few days” following the viral demo. Tesla has similarly pivoted away from motion-capture suits toward video-based training for its Optimus robot, while other robotics labs are pursuing world models to access higher-quality training data.
Key Quotes
Essentially, the world model does the same thing as the operator would do. The big unlock is essentially now that intelligence scales with the number of deployed robots, instead of the number of operators you have gathering data.
1X CEO Bernt Børnich explained how the new AI world model fundamentally changes the economics and scalability of robot training, eliminating the linear relationship between human operators and robot deployment that previously constrained the industry.
I think sometime in 2026, we will be able to ship you something that is fully autonomous out of the box and does not actually require any human intervention except for yourself.
Børnich provided a concrete timeline for when Neo robots will operate completely independently without teleoperation, signaling confidence in the rapid advancement of the company’s AI capabilities and setting expectations for the consumer robotics market.
Task execution may not always be perfect, and it may struggle at hard tasks, but it will learn over time.
A 1X spokesperson acknowledged the current limitations of autonomous operation while emphasizing the continuous learning capability enabled by the world model, managing customer expectations while highlighting the system’s improvement trajectory.
Our Take
1X’s announcement signals that AI is now sophisticated enough to train AI, creating a recursive improvement loop that could dramatically accelerate robotics development. This mirrors the broader AI industry trend toward self-improving systems and reduced human dependency in the training pipeline. The $20,000 price point positions Neo competitively against potential labor costs, though the initial reliance on teleoperation reveals the technology isn’t quite ready for prime time. The real test will be whether world models can generate sufficiently diverse, high-quality training data without human oversight—a challenge that has plagued autonomous systems across domains. If successful, 1X’s approach could establish the template for the entire humanoid robotics industry, making teleoperation a temporary bootstrapping phase rather than a permanent operational requirement. The convergence of OpenAI’s backing, Samsung’s manufacturing expertise, and advanced world models positions 1X as a serious Tesla competitor in the emerging home robotics market.
Why This Matters
This development represents a critical inflection point in AI and robotics convergence, demonstrating how advanced AI models can replace human labor even in the training process itself. The shift from teleoperation to autonomous learning through world models addresses a major bottleneck in scaling humanoid robot deployment—the need for human operators grows linearly with robot numbers, limiting scalability and increasing costs.
The broader implications extend beyond robotics to AI training generally, as companies seek more efficient data collection methods. This mirrors trends in large language model development, where synthetic data and self-supervised learning increasingly supplement human-generated training data. For workers, this represents another category of AI-adjacent jobs potentially being automated, following concerns about AI’s impact on creative, analytical, and now even physical training roles.
The race toward autonomous humanoid robots is accelerating, with 1X’s timeline suggesting commercially viable, fully autonomous home robots could arrive within two years. This has profound implications for domestic labor, elderly care, and household economics, potentially creating a multi-billion dollar market while raising questions about privacy, safety, and societal adaptation to robots in intimate home environments.
Related Stories
- The Dangers of AI Labor Displacement
- The Future of Work in an AI World
- Business Leaders Share Top 3 AI Workforce Predictions for 2025
- AI’s Role in Tech Hiring Freeze: White-Collar Job Market Slump
- Big Tech’s 2025 AI Plans: Meta, Apple, Tesla, Google Unveil Roadmap
Source: https://www.businessinsider.com/1x-humanoid-robot-training-humans-world-models-optimus-rival-2026-1