Why is it so hard to tell a robot what to do?

A new robotics company called Enigma has just raised 71 million dollars to solve this disconnect. Rather than building a robot and hoping people learn how to use its complex software, Enigma is running a massive public experiment. They have placed over 100 robots in warehouses, which are performing tasks ranging from drawing pictures to moving lab equipment. They are inviting people around the world to interact with these robots online, testing different ways of giving commands—like tapping, dragging objects on a screen, or using voice—to see which methods feel the most natural for humans.

When you ask a robot to perform a task, it doesn't just listen to your words; it has to process that information into physical motor commands. Most current approaches involve foundation models—a massive, versatile digital brain trained on vast amounts of data—to help the robot recognize objects and plan actions. However, these models still struggle with the gap between human intention and machine motion. The researchers at Enigma are looking for an interface that functions like a volume knob in a car. When you want your music louder, you don't calculate the specific decibel increase; you just turn a dial until it feels right. They want to turn human-robot communication into an similarly intuitive experience, essentially using the feedback from their public tests to teach their AI better ways to interpret human intent directly into movement.

If we can make robots as easy to direct as a common household appliance, it moves us toward a future where machines handle chores or assist in specialized fields like healthcare and logistics without needing a team of programmers to supervise them. Right now, the bottleneck isn't just the robot's physical ability to move, but our ability to guide it without feeling like we need a manual. The company that figures out that intuitive bridge could change how we interact with technology at home and at work.