It has been said that R2-D2, the squat blue-and-white robot from the Star Wars universe, gets his charm by being the perfect amount of robot—only a little anthropomorphized, but clearly more machine than human.

“Even R2-D2’s ‘voice’ avoided the uncanny valley,” wrote Clive Thomspon for Smithsonian magazine last year. “It wasn’t a voice at all but bleeps and bloops.”

The sweet spot that R2-D2 occupies might be described as: human enough to be trusted, robot enough to be helpful. And the droid’s enduring appeal has, in the 38 years since R2-D2 appeared in the original Star Wars film, extended beyond the franchise. R2-D2 helped establish an emerging cultural attitude toward robots: Machines could have personality, they could be likable. “Robby the Robot in ‘Forbidden Planet’ started to crack the friendliness barrier,” the roboticist Jerome Hamlin told The New York Times in 1984. “Then R2D2 smashed it.”

Artoo, as R2-D2 is sometimes affectionately called, has also shaped the popular perception of what a robot is and does. Robots aren’t just human replacers; they can be our helpers, and our friends. R2’s influence is visible in today’s actual robots: Roombas, coffee makers, HitchBot, Google's self-driving cars, and, more subtly, the devices most Americans carry with them pretty much all the time: smartphones.

R2-D2 was, in a way, the original mobile device—something of a hybrid between the analog world’s Swiss Army Knife and the digital world’s web-powered iPhone. In Artoo’s suite of apps: the ability to project pre-recorded holograms, conceal lightsabers, store data, perform technical maintenance work, and so on. Like an iPhone, R2-D2 is a mulitool and also a trusted companion.

Which maybe sounds like a weird thing to suggest about a cellphone. But today’s do-everything Internet-connected devices are startlingly intimate. Many people treat their phones as extensions of themselves: always aware of where they are, ever within reach, even sleeping with them tucked beneath pillows. For as much as a smartphone is a portal to the wider universe, and a way of connecting with other people, it is also a device that contains several private spaces—where people ask embarrassing questions of Google, record fleeting thoughts and grocery lists, and take selfies that will never be shared. It isn’t Siri, or any other anthropomorphized digital assistant that makes a cellphone seem so personal; it’s how much of themselves that people pour into their devices—a different way of projecting humanness onto a machine.

It’s understandable, the idea that robots have to be non-threatening in order for humans to accept them, or at least like them. But to be non-threatening does not necessarily mean human-like. R2-D2, after all, is far more popular than his humanoid sidekick, C-3PO.

The thing is, neither R2 nor C-3PO reflect what most actual robots are like in our world. The rise of robots is already occurring all around us, and in many ways undetected. Algorithmic bots crawl human-made websites, while assembly-line robots make candy bars and television sets. Robots already work in customer service and in banks. They pick strawberries in orchards and retrieve goods in Amazon warehouses. They predict the weather, take inventory, and help human surgeons perform complex operations. In the near future, the writer Kevin Kelly predicts, people will be paid based on how well they work with robots: “Ninety percent of your coworkers,” he wrote for Wired, “will be unseen machines.”

“In fact, as mobile peripheral devices become more and more common, it may be increasingly difficult to say exactly what a robot is,” the Microsoft co-founder Bill Gates wrote for Scientific American in 2007. “Because the new machines will be so specialized and ubiquitous—and look so little like the two-legged automatons of science fiction—we probably will not even call them robots.”

R2-D2 may have helped train humans to see robots as endearing. But people often like robots best when we they don’t notice them at all.