Self-driving cars are often held up as the most tangible example of the coming age of AI. Humans and computers will quite literally share the road and algorithms will be compelled to make life-or-death decisions in the blink of an eye. For those reasons, here at Unanimous AI we’ve partnered with the MIT Media Lab to uncover the optimal strategies for programming AI decision-making, and have written often about the public’s perception of autonomous vehicles.
Recently, Waymo has obtained permission to operate fully driverless cars in California. Until now, self-driving cars typically retained a human driver in the front seat for oversight, even if the vehicle itself was fully capable of autonomous operation. But now Waymo, the self-driving unit of Google parent Alphabet, will be able to test its cars on public roads with speeds of up to 65mph. With this new development in mind, researchers at Unanimous AI convened a swarm of American consumers to generate AI-optimized insight into their perspective on the future of driverless cars.
First, Unanimous researchers wanted to know how likely the Swarm AI system believed that a person like themselves would be willing to ride in a driverless car. Like many new technologies, what at first glance seems completely foreign eventually becomes routine, but at this moment in time, the swarm of potential passengers converged on the answer that less than a third of would-be passengers would willingly ride in a completely driverless car?
Inherent to understanding this customer reluctance is the relationships between humans and technology. It’s one thing to use an AI like Siri or Alexa to remind you to take the cookies out of the oven, it’s another thing entirely to trust an AI to pilot you safely down the road at 65mph. So, despite the evidence that self-driving cars are already safer than their human counterparts, we tend to hold AIs to a much higher standard. That might be why the Swarm AI system’s response when tasked with ranking the trustworthiness of a driverless car versus the equivalent Uber or Lyft driver turned out to be a clear victory for humans.
The best ride is the ride in which nothing happens, and the passenger is delivered to their location without incident or inconvenience. But, accidents do happen, and researchers at Unanimous AI wanted to know, all things being equal, if passengers would be more annoyed by a minor accident in a car driven by a human, or one with no driver at all. The presented scenario posited a passenger who often uses ride-sharing programs, with an equal propensity to ride in driven and driverless automobiles.
Here you can see a swarm divided. Unlike the previous response, where the AI was able to quickly converge on the answer, on this question, the group saw strong support for both driving options, and ultimately landed on the human driver, albeit by a small margin.
These results suggest that most potential passengers remain apprehensive about the coming age of the driverless car. Irregardless of whether or not a person likes to drive, the notion of completely abdicating the responsibility to an AI algorithm still gives most people pause. This reluctance might be due to the revelation that this swarm of potential passengers still trusts a human’s ability to safely deliver them to their destination, and is bolstered by the fact that they would a human more responsible than an AI for any minor accident that occurred along the way. The research suggests that this swarm doesn’t yet believe a driverless car is capable of accepting responsibility for its actions, which makes them a less attractive option.
About Unanimous AI:
Unanimous AI is a Silicon Valley company that has pioneered Swarm AI® technology, a new form of AI that combines real-time human insights and AI algorithms modeled after natural swarms. In 2018, Swarm AI technology won “AI Innovation of the Year” at the SXSW Innovation Awards. For more on Unanimous, visit http://unanimous.ai
Want to learn more about our Swarm AI technology? Check out our TED talk below…