In the realm of AI, where poems, essays, and books are no match for its capabilities, researchers from TU Delft and EPFL have embarked on a quest to explore the boundaries further. They wondered, could the open AI platform ChatGPT be used to design a robot? This inquiry led them to delve into the potentials and risks of incorporating AI into the design process.The researchers' inquiry began with the question of humanity's greatest future challenges. With ChatGPT as their conversational partner, they focused on the food supply challenge. Eventually, they arrived at the idea of creating a tomato-harvesting robot, with ChatGPT guiding the design process.ChatGPT proved invaluable during the conceptual phase, expanding the designers' knowledge beyond their expertise. It even provided insights on the most economically valuable crop to automate. As the implementation phase unfolded, ChatGPT continued to offer helpful suggestions, such as using silicone or rubber for the gripper to prevent tomato damage and employing Dynamixel motors for optimal robot movement. The collaboration between humans and AI resulted in a robotic arm capable of efficiently harvesting tomatoes.The researchers found the collaborative design process to be positive and enriching. However, they noted a shift in their roles as engineers, with a greater emphasis on technical tasks. In their study published in Nature Machine Intelligence, they explored the extent of cooperation between humans and Large Language Models (LLMs) like ChatGPT. In extreme scenarios, AI could provide all the input, while humans blindly follow. Here, LLMs act as researchers and engineers, while humans serve as managers, defining design objectives.Yet, relying solely on LLMs for design decisions is not without risks. Unverified or unvalidated outputs from AI can lead to misinformation and bias in the field of robotics. Issues such as plagiarism, traceability, and intellectual property also arise when working with LLMs. Therefore, caution must be exercised, and human judgment remains essential.The researchers plan to further explore the potential of LLMs in robot design and continue their research on robotics using the tomato-harvesting robot. Their focus extends to examining the autonomy of AI in designing its own bodies. They raise thought-provoking questions about how LLMs can assist robot developers without hindering creativity and innovation, critical for meeting the challenges of the 21st century.As the boundaries of AI expand, it is crucial to strike a balance between AI's assistance and human ingenuity. The future of robotics lies in harnessing the power of LLMs while ensuring they augment rather than limit human creativity and innovation.