The Hidden Labor Problem Inside the Humanoid Robot Boom

Key points

• Humanoid robotics companies raised billions of dollars over the past 18 months, much of it funding large-scale human teleoperation programs.
• Teleoperation datasets are more than 100,000 times smaller than the datasets used to train today's language models.
• Workers across China, India, Europe and the US are being hired to film household tasks and operate robots remotely, forming a commercial data-supply industry.
• Nikita Rudin, co-founder and CEO of Flexion, argues that reinforcement learning in simulation is the only approach that can break the dependency on human demonstration data.
• Flexion raised 50 million dollars from DST Global and NVentures to build what Rudin calls a general-purpose "brain" for humanoid robots.

The pitch for humanoid robots has always been straightforward: humans won't be able to fill certain jobs in the future, so robots will step in. A pointed opinion piece published by The Robot Report this week raises an uncomfortable counter-claim. The robots, it turns out, may need a permanent supply of human workers just to function.

The issue is teleoperation, where a human operator controls a robot through a remote device the way you might steer a toy car with a controller. Every movement gets recorded. Then the robot studies that footage and tries to copy it.

Teaching a robot this way sounds sensible. The problem, according to Nikita Rudin, who completed his PhD at ETH Zurich's Robotic Systems Lab and worked on Isaac Gym and Isaac Lab simulation tools at Nvidia, is that it scales terribly.

A shelf moves. A door handle is slightly different. A new box shape arrives on a factory floor. Each tiny change means someone has to film the task again from scratch, and the workforce generating demonstrations can never keep pace with a world that doesn't hold still.

Data quality compounds the problem. Operators controlling a robot remotely can't feel what the robot is touching or judge distance accurately, so they move slowly and overcorrect. The robot then learns from footage of someone struggling with a controller, which is exactly what it practises.

Will more human operators actually solve this?

No, and that is the core of Rudin's argument. He draws a comparison to the early days of large language models, the AI technology behind chatbots such as ChatGPT. Those early models were trained on enormous amounts of text and could loosely copy the style of Shakespeare, but couldn't reason. The real leap came through reinforcement learning, a training method where software learns by trying, failing and adjusting across millions of attempts, with no human directing each step.

Rudin argues humanoid robotics is stuck at that earlier, imitation-only stage. More teleoperation data means more imitation at massive cost.

The alternative he describes is reinforcement learning inside computer simulations. A robot's software can fail millions of times in a virtual world, reset instantly, and run in parallel across many computers at once. More computing power directly means faster progress. That is a very different equation from hiring workers in lower-wage countries to film themselves doing laundry.

We first covered Flexion on 19 July 2026, when Rudin's funding round was announced; the sim-to-real argument he's now making publicly sharpens what was then a product claim into a direct challenge to the industry's dominant training method.

For the people doing that teleoperation work today, and for investors funding these companies, Rudin's challenge is direct: if robot autonomy is genuinely the goal, the data should show that human involvement is decreasing over time. If it isn't, teleoperation has stopped being a bridge and quietly become the permanent method. That's the question worth asking every robotics company that's still hiring operators at scale.