Key Takeaways:
- The robotics industry is shifting from a paradigm of pre-programming machines with complex mathematics to one where robots learn directly from data and experience
- Robots are being developed to "think" for themselves, enabling them to perform tasks that were previously difficult or impossible, such as opening doors or climbing stairs
- The goal is to create "generally intelligent software" that can act as a brain for any robot hardware, reducing costs by an order of magnitude
- The lack of data for robot physical interactions is a major hurdle, but companies like Skild are working to create a "data flywheel" to overcome this challenge
- Robots are expected to proliferate in industrial settings and semi-structured environments before entering the home environment
Introduction to the Shift in Robotics
The robotics industry is undergoing a significant transformation, with a shift from pre-programming machines with complex mathematics to enabling robots to learn directly from data and experience. According to industry executives at the Fortune Brainstorm AI conference, the true revolution in robotics is not about physical agility, but about the ability of robots to "think" for themselves. As Skild AI CEO Deepak Pathak noted, "The change is things in robotics used to be driven more by human intelligence… What has now changed is that these models or these robots can now can learn from data." This new approach is enabling robots to perform tasks that were previously difficult or impossible, such as opening doors or climbing stairs.
The Old Paradigm of Robotics
For the past 70 years, robotics relied on a specific paradigm: intelligent humans pre-programming machines with complex mathematics to execute specific tasks. This approach is now obsolete, argued Sequoia Capital partner Stephanie Zhan and Skild AI CEO Deepak Pathak. The industry is undergoing a massive shift where robots, much like the Large Language Models (LLMs) behind tools like ChatGPT, are learning directly from data and experience rather than following rigid code. As Pathak explained, "What has now changed is that these models or these robots can now can learn from data." This shift is enabling robots to become more autonomous and adaptable, and to perform tasks that require a high degree of flexibility and problem-solving.
The Challenge of Physical Intelligence
The real challenge in robotics lies in interaction with the chaotic real world. Climbing stairs or picking up a glass requires a robot to continuously use vision to correct its movements in response to a changing environment. This "sensory motor common sense" is the root of human general intelligence, and it is the barrier that new "brain" software is attempting to break. As Pathak noted, "It’s actually a lot easier to program a robot to do a backflip than it is to get them to climb stairs." This is because backflips require controlling the robot’s own body in free space, a physics problem that computers have been good at solving for decades. In contrast, climbing stairs requires a robot to interact with a dynamic and unpredictable environment, which is a much more complex task.
The Market Opportunity
Investors and executives see this as a market opportunity comparable to the recent explosion in generative AI. Zhan noted that just as OpenAI unlocked the market for digital knowledge work, companies like Pathak’s Skild are aiming to unlock the market for all physical labor. The goal is to create "generally intelligent software" that can act as a brain for any robot hardware, reducing costs by an order of magnitude. As Zhan explained, "The goal is to create a brain that can be applied to any robot hardware, so that you can have a robot that can do anything." This would enable robots to perform a wide range of tasks, from manufacturing and logistics to healthcare and hospitality.
The Lack of Data
Unlike the software world, however, robotics faces a unique hurdle: a lack of data. While LLMs were trained on the entire internet, there is no equivalent database for robot physical interactions. Pathak argued that the company that deploys first will win by creating a "data flywheel," in which field robots generate the data needed to make the system smarter. As Pathak noted, "The company that deploys first will win, because they will be able to create a data flywheel that will make their system smarter and smarter over time." This data flywheel will enable robots to learn from their experiences and adapt to new situations, making them more autonomous and effective.
The Timeline for Robot Adoption
For consumers wondering when a robot will be doing their laundry, the timeline remains staged. Pathak and Zhan predicted that robots will first proliferate in industrial settings and "semi-structured" environments like hotels and hospitals before entering the more chaotic environment of a private home. As Pathak explained, "Robots will first be adopted in industrial settings and semi-structured environments, and then they will move into the home environment." This is because industrial and semi-structured environments are more predictable and controlled, making it easier for robots to operate effectively.
The Future of Work
Despite fears of job displacement, Pathak and Zhan argued that the technology is necessary to address the "Three S’s" of the future: Safety, Shortages, and Social evolution. Robots are poised to take over jobs that currently force humans to risk their lives or health. Furthermore, with millions of job openings currently unfilled due to labor shortages, robots could fill the gap in essential blue-collar work. As Zhan noted, "Robots can help address the shortage of workers in certain industries, and they can also help to improve safety and reduce the risk of injury." Ultimately, the hope is for a social shift where dangerous or drudge work becomes optional, allowing humans to focus on tasks they enjoy.
https://fortune.com/2026/01/06/robots-advancing-from-backflips-to-opening-door-handles-skild-sequoia/