In robotics, there may be its own ChatGPT moment. What does that mean?

What the “ChatGPT moment” means for robotics

Before the advent of GPT-3, companies built their own text-processing models by training them on massive amounts of specialized data. With the emergence of universal language models, developers gained the ability to use a ready-made foundation and adapt it to their specific tasks.

According to General Intuition founder Pim de Witte, a similar transformation is awaiting robotics. Instead of training artificial intelligence separately for each robot, companies will be able to use a single foundational model that already understands how to interact with the surrounding world.

Why modern robotics requires too much data

Today, most robotics companies collect hundreds of thousands or even millions of hours of data from cameras, sensors, and other devices. This data is necessary for robots to learn even relatively simple actions.

This approach requires significant time and financial resources, and each new robot model effectively starts its training from scratch.

What General Intuition offers

Instead of training on real robots, the company decided to use another data source: video games.

Developers trained a foundational AI model on millions of hours of gameplay, where the artificial intelligence analyzed human actions: which buttons were pressed, how the character moved, and how decisions were made in space.

According to the team, such data helps form an understanding of movement, time, and interaction with the environment.

How the model works in practice

After training, the model was tested on a quadruped robot.

Adapting to the real device required only eight minutes of real-world data. After that, the robot was able to navigate an office using only footage from a single front-facing camera, without additional sensors.

According to developers, the model performed successfully even when people were moving around the room and the environment was constantly changing.

Why this is important for the industry

If this approach proves successful, developers will no longer need to collect massive datasets for every new device.

Companies will be able to use a universal AI model as a foundation and only slightly adapt it for their own robots. This will significantly reduce development costs and accelerate the time to market for new products.

A similar scenario has already occurred in generative artificial intelligence, when developers stopped creating their own models and began using GPT, Claude, Llama, and other foundational models.

Why investors are betting on this approach

In June 2026, General Intuition raised $320 million in funding at a valuation of $2.3 billion.

The company does not plan to manufacture robots itself. Its goal is to create a universal AI platform for physical artificial intelligence that robotics manufacturers worldwide can use.

If this strategy proves successful, the market could receive a ChatGPT-like equivalent for robots—a universal model that will serve as the foundation for creating a wide variety of autonomous machines, from industrial robots to autonomous vehicles.