Skip to main content

Galaxy Robot Founder: Embodied AI's ChatGPT Moment by 2028

Embodied AI's Big Leap: A ChatGPT-Like Breakthrough on the Horizon

At the 2026 World Artificial Intelligence Conference, Wang He, founder and CTO of Galaxy General Robotics, made a bold prediction: embodied intelligence—AI that can interact with the physical world—will reach a turning point by 2028, comparable to the moment ChatGPT revolutionized conversational AI.

The 70-80% Success Rate Goal

Wang He explained that future embodied intelligence foundation models, trained on massive datasets, could achieve a 70-80% success rate on tasks they've never been specifically trained for. That's a game-changer. Think of it like a robot that can figure out how to fold laundry or assemble furniture without step-by-step instructions—just by understanding the general concept.

"This milestone will be on par with early digital models' conversational abilities," Wang said. To get there, he emphasized the need for powerful pre-training models and efficient post-training paradigms. The industry has already seen rapid evolution in large models for embodied intelligence over the past two years, and Wang's prediction only fueled more discussion among attendees.

From Household Chores to Complex Decisions

So what does this mean for everyday life? Wang envisions robots that not only handle simple chores like cleaning or cooking but also participate in complex decision-making and analysis. Imagine a robot that helps you plan a renovation, manage your schedule, or even assist in medical diagnostics. The possibilities are vast, and the transformation across industries could be profound.

"Embodied intelligence will bring new experiences and possibilities to human production and life," Wang noted. It's not just about automation—it's about enhancing our quality of life and work efficiency.

The Road Ahead

Wang's confidence stems from the accelerating pace of data accumulation and model convergence. He believes the industry's progress will be rapid, driven by both technological advances and growing investment. While some skeptics question the timeline, Wang's vision has ignited excitement and debate, pushing researchers and professionals to think bigger.

Looking forward, embodied intelligence could become as ubiquitous as smartphones, changing how we interact with machines and the world around us. Wang's prediction serves as both a roadmap and a rallying cry for the AI community.

Key Points

  • Timeline: Wang He predicts embodied intelligence will achieve a ChatGPT-like breakthrough by 2028.
  • Performance: Future models could handle 70-80% of untrained tasks, matching early conversational AI.
  • Applications: From household chores to complex decision-making, robots will transform daily life and industries.
  • Drivers: Rapid data accumulation, model convergence, and efficient training paradigms are key to progress.
  • Impact: The prediction has sparked lively debate and high expectations for the future of embodied AI.