In recent years, we’ve been constantly bombarded with news about the launch of new robots, whose mobility and intelligence capabilities far surpass any previous versions. These reports often include headlines—sometimes alarming—about major societal changes. However, these changes have yet to materialize, either due to a lack of accuracy or an excess of enthusiasm.
At a first glance, we find news related to job losses caused by the presence of robots in the workplace. Without considering other aspects, these claims can easily be refuted with official data, as countries with the highest number of industrial robots per employee, such as South Korea or Singapore, have unemployment rates below 3%.
At the same time, the excessive enthusiasm in the headlines contributes to generating expectations that are difficult to meet in the short term. In fact, if even a small percentage of these reports were accurate, we would already be surrounded by robots in our homes and interacting with them daily. To analyze why this hasn’t happened (beyond robotic vacuum cleaners, whose classification as robots is somewhat debatable), we need to look at several features that current robots have yet to combine: autonomy, the ability to solve general problems, and affordability.
Robot autonomy has improved greatly, thanks to the development of new batteries—driven by the popularity of electric vehicles—and advancements in legged robots. The latter are much more versatile than wheeled robots, but their movements tend to be slow and unstable, as the technology has not yet matured enough for widespread use. Nonetheless, we already see solutions like Spot, from Boston Dynamics, capable of overcoming significant mobility challenges. Its price is still too high for individual consumers, but it is already being used in professional environments.
The second point to consider is the ability to solve different kinds of problems—a necessary step toward general-purpose robots capable of performing various tasks, such as ironing, dusting, or putting away groceries. Assuming robots equipped with powerful arms and hands, the main challenge lies in the robot’s intelligence, including its adaptability after deployment. Recent advances in AI—including Large Language Models (LLMs), the technology behind ChatGPT—have opened new avenues for exploration in the coming years, enriching a robot’s understanding of its environment and improving decision-making. The potential LLMs offer for robotics is enormous, and we hope their effective integration will help realize general-purpose robots. However, this integration is not without risks, such as potential biases (racial or gender) or the lack of explainability in decision-making—issues that, in a physical platform, could have much more serious consequences.
Engineer and PhD in Computer Science (University of Castilla-La Mancha). Associate Professor at the School of Computer Engineering in Albacete and researcher in the Intelligent Systems and Data Mining group at the Albacete Research Institute of Informatics (i3A).