Decades before Google started outfitting Lexus SUVs with sensors and self-driving software, the driverless car du jour was an otherwise ordinary-looking Pontiac.

In the 1920s and 1930s, a driverless car was more commonly known as a “phantom auto,” and demonstrations of the technology drew thousands of spectators in cities across the United States.

These cars weren’t computer-driven, as they are today, but remote controlled. It’s not clear from newspaper archives and other written accounts how many were ever in existence, anyway, though we do get a general sense of how they worked: The person operating the car would often follow in a second vehicle some distance behind; or, in at least one case, in a low-flying airplane, according to a 1932 account in the Times Recorder of Zanesville, Ohio.

The movie star Helen Twelvetrees with her 1935 Pontiac, the same model that was remote-controlled during “phantom auto” demonstrations. (Mitchell Library, State Library of New South Wales)

The car, which the newspaper called “one of the most amazing products of modern science,” could be operated from as far as five miles away, its inventor said. “It sounds unbelievable but it is true that the driverless car will travel about the city through the heaviest traffic—starting, stopping, sounding its horn, turning right or left, making U-turns and circles, and proceeding thus as though there were an invisible driver at the wheel,” the Times Recorder reported. To catch a glimpse of the phantom auto was to see “one of the most spectacular street events possible.”

The Greeley Daily-Tribune, 1935 (

The cars were driverless, in that they didn’t have a human behind the wheel, but they weren’t self-driving per se. This didn’t stop them from captivating the public. “From 1931 to 1949, [the radio engineer J.J.] Lynch gave demonstrations of the remote-controlled vehicle in 37 of the 48 US states,” wrote Fabian Kröger in a sprawling report on the technical, legal, and social aspects of self-driving cars. “He manipulated the brakes, steering wheel and horn of the vehicle driving in front of him with the aid of a morse key. A spherical antenna received the code, although there are also reports of a wire between the vehicles.”

At the same time, Kröger points out, driverless vehicles were cemented in the popular imagination through their depiction in fantasy novels like Werner Illing’s Utopolis, which features a self-steering car. “The most wonderfullest thing about it was that the car … behaved as if it had learnt all possible traffic rules by heart,” Illing wrote. In 1935, the writer David H. Keller wrote about a driverless car that was voice activated in The Living Machine:

Old people began to cross the continent in their own cars. Young people found the driverless car admirable for petting. The blind for the first time were safe. Parents found they could more safely send their children to school in the new car than in the old cars with a chauffeur.

Decades later, KITT, the sentient automobile from Knight Rider, became the best known fictional car you could talk to. Incidentally, it was a Pontiac; just like the actual phantom autos of the 1930s.

But when the concept of a driverless car first emerged, the thing people focused on most was the promise of improved safety at a time when vehicles were deadly—which is still the great promise of self-driving cars today. In fact, the driverless car’s first film appearance was in a 1935 educational road-safety movie commissioned by General Motors. But remote-controlled vehicles seemed to exist to make a point more than anything else: They were a way of demonstrating how safe cars ought to be, but still just a novelty and not a prototype for the near-future. In the film, The Safest Place, the driverless car’s traffic record is “exemplary,”Kröger says:

The vehicle always stays in its lane, never forgets to signal when turning, obeys all stop signs and never overtakes on dangerous corners. Lynch had given similar reasons for campaigning for safety with driverless vehicles. ... In ironic fashion, the film points here to the contradiction between safety and freedom: Is the car only safe when empty?

Such public education was sorely needed. In the 1920s and 1930s, automobiles were responsible for so many deaths that public officials began to debate whether cars were inherently evil. Cities across the country held safety parades in which disfigured survivors of accidents sat waving from automobiles—and thousands of children dressed up as the ghosts of the deceased. And though driverless cars—often referred to in the newspapers as “magic cars”—attracted public attention as a a technological marvel, public demonstrations were mostly intended to underscore the need for improved automotive safety. (It would take decades more for American attitudes toward automotive safety to shift, and decades more for people to begin to think of self-driving cars as a technology that might be adopted widely.)

“Regular safety lectures leave a sour taste in everybody’s mouth, especially when you start telling another fellow about his shortcomings as a driver,” J.J. Lynch, the phantom-auto operator, told The Daily Times-News of Burlington, North Carolina, in 1937. “But when you give them this kind of demonstration and talk safety at the same time, they listen to you and become interested.”

Driverless cars never ran yellow lights or veered into the wrong lane, Lynch added, and “that’s more than a lot of cars with drivers in them do.”