In fiction it is not strange to find the figure of the assistant robot. This is usually a fully self-aware humanoid robot, a type of sapiens robot programmed to collaborate on certain tasks with humans. Some of the best known are C-3PO, from Star Wars , and Baymax, from Big Hero 6 .

In these stories there is usually a common element: all the characters in the work normally accept this figure. For example, no one feels uncomfortable next to Bender in Futurama because he is a robot. The integration of robots into society is complete and they are completely accepted. But social approval of a completely autonomous robot adapted to the social environment would not be so easy to achieve in the real world.

Let’s put it in the case that fiction becomes reality and the first completely autonomous robot adapted to the social environment is ready to be inserted in cities. How would it be received by users? What design aspects are key to improving that reception?

Natural communication

A crucial aspect for the proliferation of these robots on a daily basis to be well received is the way we will have to interact with them. Or what is the same, the user experience . In fiction, with Bender or C3PO, communication with automatons barely differs from that established between humans. They completely understand what they are told and, even more importantly, they communicate in a natural way. This implies that you do not have to use any type of interface or artifice to chat with the robot: we simply speak and it captures the message as if it were another person.

This understanding is not that far off in reality. We have probably all used similar technologies: Cortana, Siri and Alexa, the most popular virtual assistants, are capable of verbal communication and maintaining a “conversation” without problems. Now, they have a limitation in their understanding, since they are programmed to respond to known basic guidelines .

For example, they perfectly interpret messages such as “Alexa turn on the lights in the house” or “Siri, search the internet for the latest song by X artist.” But if we go further, their performance is greatly reduced, or they even stop understanding us.

When ChatGPT can talk

This technology, which has quickly been inserted into homes without social rejection, will experience a great advance driven by artificial intelligence. AI makes it possible to design and build models that understand natural language, such as ChatGPT or Google Gemini , apparently capable of understanding us almost perfectly when we ask them questions or simply chat.

However, today communication with these models takes place through written chat and not voice. Having ChatGpt speak will make communication easier and flow without restrictions of any kind.

Another key to human communication is empathy . It doesn’t have to be at an advanced level, but the ability, as a conversation develops, to change our message or attitude depending on the reactions the recipient is having makes communication much more natural. For example, if we realize that the message we are giving to someone is affecting them, we can soften it or change the words we use.

Show feelings

Rosalind Picard, professor and director of a research group at MIT, assures in her book Affective Computing that to achieve a natural interaction with machines, and for them to be truly intelligent, they would have to be able to recognize and understand feelings, and even show your own feelings.

This problem is still open, and it is a fairly active branch of research in the field of biometrics, where researchers focus on trying to recognize emotions through gestures, voice tones and any other information that accompanies words .

Although all this may sound like fantasy, in the real world there are already some social robots operating. And although they do it on a smaller scale than in fiction, they work side by side with people and, little by little, they enter society.

An example of this is the social robot Pepper, active to assist children with hearing deficiencies . Another is Stevie, a robot that assisted in a nursing home , with very good reception among the residents.

The most important thing for the insertion of that “perfect” humanoid robot is to get it to act completely naturally with us , to be able to understand us when we speak and respond appropriately, to see us angry and change its attitude, and even to show happiness and others. human feelings .

Author Bios: José Ignacio Salas Cáceres is a Researcher in Computational Engineering and Javier Lorenzo Navarro is an Associate Professor both at the University of Las Palmas de Gran Canaria