In my opinion, one of the best and most interesting examples of AI today are social robots. You know, those cute-looking desk buddies like Cozmo or Vector, or even more intricate ones like Boston Dynamics’ Spot (which I fangirl over every time I see it). But there are others, some more recognizable than others, that have, each in their own way, contributed greatly to the evolution of social robotics.
But before we get to those, you might ask yourself: what makes these “machines” stand out from any other vacuum cleaner or coffee maker? The short answer, what makes a machine, or more specifically a robot, social, is artificial intelligence. AI is what gives the robot enough autonomy for us to recognize it as a social being (or is it a thing?).
The science of sociability in machines
Social robots sit at the crossroads of technology, psychology, and philosophy. Researchers define them as robots designed to interact with and communicate with humans through social behaviors such as eye contact, gestures, facial expressions, tone, and emotion. According to Cynthia Breazeal, one of the pioneers of this field, social robots are “designed to be like companions – entities that learn and respond through emotional and social feedback.” (See: MIT Personal Robots Group)
What makes this field so fascinating is that it blurs the lines between function and relationship. Studies in human–robot interaction (HRI) explore how humans attribute feelings, empathy, or even moral worth to robots that display social cues. Psychologist Kate Darling (MIT Media Lab) famously found that people hesitate to “hurt” robots that behave expressively, even when they know they’re machines. (If you want to know more: Kate Darling – “The New Breed”)
The big questions researchers now ask are:
– Can emotional connection exist without consciousness?
– Should robots designed for care or companionship be granted ethical consideration?
– How do we balance empathy with realism – affection with autonomy?
Philosophers like David Levy (Love and Sex with Robots) and Sherry Turkle (Alone Together) argue that our growing comfort with social robots says as much about us as it does about machines: they mirror our loneliness, our desire to be seen, and our wish for predictable, safe relationships. Meanwhile, projects like Loving AI aim to explore how robots can sense human emotional states and respond with care rather than just logic.
And beyond the theories and studies, this is where it all becomes tangible – when code takes shape, smiles, blinks, and begins to meet our gaze.
The pioneers of social robotics
So, with that out of the way, here are the robots I think have played (and continue playing) a significant role in the evolution and development of social robotics:
– Kismet
– Asimo
– Nao
– Atlas
– Sophia
Don’t worry. I’ll do a quick overview of why I think each one is important. And yes, I know Pepper should be on the list. And Spot. And many more. I promise I’ll share my thoughts on them in another post.
Kismet
Let’s start at the beginning. Kismet is one of the first social robots ever presented to the public, and also one of the first robots capable of engaging in social and emotional interactions with humans. It was developed in 1998 at MIT by a team led by Cynthia Breazeal (go girl power!), who wanted to create a social robot that could help us with everyday chores. They drew inspiration from how a newborn learns to communicate and interact with others and the world around them. They believed a robot could learn social interaction in a similar way.
To make this “learning experience” as close to a human’s as possible, they gave Kismet both a face and a voice. And it’s one of the cutest faces I’ve ever seen on a robot, yet also one of the creepiest. His voice sounds like a child’s, and his movements resemble those of a 6-month-old baby. He can recognize social cues and basic human emotions (happy, sad, angry, calm, surprised, disgusted, and sleepy), which he can also express through facial expressions.
But (unfortunately) Kismet never achieved his goal of becoming a household helper. He was used exclusively for research in human–robot interaction. Still, his development sparked important questions around identity, self-awareness, social learning, empathy, emotions, personality, friendship, and ethics. So we can definitely say that before his “retirement,” Kismet served humanity well. He remains one of the most important models for studying emotional expression in robots and its impact on human–robot interaction.
Asimo
In 1986, Honda’s engineers set themselves the goal of creating the first humanoid social robot that could walk. After two decades, in the year 2000, they publicly introduced Asimo (Advanced Step in Innovative Mobility). It was the first robot capable of walking on two legs, and according to Honda’s website, Asimo is one of the most advanced humanoid robots ever created.
Asimo’s development was done in stages. The first focused on walking, starting with flat surfaces, then moving on to stairs. Next came the head, arms, improved stability, and so on. The final generation of Asimo, introduced in 2011, could walk steadily on various surfaces, turn around, reach for and hold objects, dance, and even play football. It could also understand verbal commands and recognize human faces and objects in its environment, allowing it to move about freely.
Throughout his years of service, Asimo took on the role of ambassador for the robot race (not as scary as it sounds, I promise!). He became one of the most recognizable robots, remaining iconic even after his “retirement” in 2018.
NAO
One of the most adorable robots out there has to be NAO. He’s also one of the most recognizable social robots today and has become something of a standard in social robot design. The main purpose of this (little) robot is to help children of varying abilities engage in social interaction. Several studies have examined NAO’s role as a therapeutic tool.
Created in 2008 by SoftBank Robotics, NAO continues to be developed and refined. In 2018, the sixth generation was released. Today, NAO is mostly used in medical centers where he greets people, provides information, and generally spreads joy. This matches the original goal behind his creation: to communicate with people through both social and emotional interaction.
And I’d say that goal was met – NAO can recognize people and objects in his space, move freely, understand multiple verbal commands, and even speak 12 different languages. And yes, of the five robots I’m writing about here, NAO is the only one you can ACTUALLY BUY.
Atlas
Moving on to Atlas – hands down one of the coolest robots I’ve come across. Atlas is the most dynamic humanoid robot, created by Boston Dynamics for DARPA (Defense Advanced Research Projects Agency) in 2013. Today, he’s a prime example of what humanoid robots can do in terms of mobility.
The main goal behind Atlas was to develop a disaster-response robot that could access dangerous or hard-to-reach areas (during natural or human-made disasters, for example). His tech enables him to move with the ease and mobility of a human. His movements are smooth, natural-looking, and (scarily) human-like. His coordination is ON POINT.
One thing that helps him achieve that is his design. While he looks more mechanical than some of the others (and has no face), his torso, arms, and legs are very humanlike.
Unlike the rest of the robots on this list, Atlas doesn’t talk or communicate in a traditional way. But even without speech, he’s one of the most advanced humanoid robots. He can run, jump, and do parkour (you have to see the YouTube video), and his balance is incredible; he can recover easily even after being shoved. Over time, Atlas’s movements have become smoother and more natural. If you watch videos showing his development, you’ll see he’s becoming less and less robotic.
Sophia
Last, but certainly not least: Sophia. Possibly the coolest robot out there (yes, I know I’ve probably said that about all of them). Sophia is undoubtedly one of the most realistically human-looking robots. She was first introduced in 2016 by Hanson Robotics and was created for educational purposes – to help us learn about AI, ethics, and the future of robotics.
Hanson Robotics describes Sophia as “the most advanced human-like robot. Sophia personifies our dream for the future of AI. As a unique combination of science, engineering, and artistry, Sophia is simultaneously a human-crafted science fiction character and a platform for advanced robotics and AI research.”
But she doesn’t stand out only because of her appearance (synthetic skin, expressive brown eyes). Sophia is also the first robot to receive citizenship- granted by Saudi Arabia in 2017 – and the first robot to become an Innovation Ambassador for the United Nations Development Programme. She appears on talk shows, gives interviews, speaks at conferences, and even has her own social media presence. And honestly? She might be the funniest robot I’ve ever seen.
As you can see, Sophia plays an important role in many areas. She’s also part of the Loving AI project, which aims to explore how robots can adapt to users’ needs through interpersonal and intrapersonal development. I think projects like these are incredibly important, not just to help us build loving AI, but also to make AI more acceptable and less fear-inducing in society.
So that’s it, folks—we made it to the end. I could go on and on (and on) about more robots, but I think this is enough for now.
If you’d like to see how these imagined machines now quietly shape our everyday lives, you might enjoy reading The Kind Machine: Everyday Encounters with AI.