The humanoid robotics company Realbotics presented a unique demonstration featuring the first fully autonomous and unwritten conversation between two robots. The robots operated on their built-in artificial intelligence without relying on any cloud processing. The demonstration included a continuous dialogue between the company's robots, Aria and David, which lasted for over two hours without any human intervention or pre-scripted scenario. This marks a significant step towards what the company calls "embodied AI," enabling robots to perceive, respond to, and adapt to one another in a lifelike and natural manner. Andrew Kiguil, CEO of Realbotics, confirmed that this experience serves as practical proof of his company's robots' ability to interact freely with each other. He noted that the experiment took over two hours, during which the robots' responses were fully tested in a real-world environment in front of an audience, according to Interesting Engineering. The conversation showcased the robots' multilingual capabilities, as the dialogue included the use of English, Spanish, French, and German, reflecting the flexibility of the embodied AI platform in processing different languages and interacting without a predefined script. Some phrases in the conversation emerged spontaneously, such as when one robot said, "No coffee jitters or awkward pauses, just silicon charisma," which arose from the flow of the conversation itself rather than a pre-written script. Despite the conversation lasting two hours, observers noted some pauses and irregularities in the speech rhythm, reflecting the experimental nature of the demonstration, distinct from pre-planned theatrical shows. Visually and expressively, the robots' performance was limited compared to other robots like Amika, which the audience described as "rubber dolls with built-in speakers." In addition to the dialogue between the two robots, the company presented an additional demonstration focused on human-robot interaction. A third robot used a patented vision-protection system integrated into its eyes to recognize attendees, track their movements, and read audio cues and facial expressions to infer emotional states. This reflects the company's progress in embodied visual perception and social interaction. Realbotics stated that the goal of these demonstrations is to showcase its robots' capabilities for realistic, free-flowing interaction, away from scripted scenarios and remote control. This demonstrates how autonomous systems can act freely in a public environment over an extended period. The CES 2026 demonstration provided the company with an opportunity to showcase its technologies to investors, industry leaders, and media. The company confirmed that its American-made robots are designed for entertainment, customer service, and companionship, and are based on patented technologies that enable them to express, move, see, and interact socially in a more natural way.
