When Rubber Meets the Road: AV as a Harbinger of AI

Over the next decade, Autonomous Vehicles (AVs) will introduce most of the world to Artificial Intelligence (AI) for the very first time. While autonomous vehicles will save thousands of lives and generate significant economic growth, they will also displace millions of jobs and spur new philosophical debates. Compared to today’s abstract applications of AI in data analysis and computer vision, trusting an artificial intelligence to merge you onto the highway is a much more tangible and visceral experience. The public’s reaction to the new AV paradigm will define how deeply AI integrates into society beyond 2032 – and the critical early success (or failure) of AV technology will pave the way for all future applications of AI.

Somebody dies in traffic every 24 seconds. Every year, Americans waste billions of hours stuck in traffic, and conservative estimates calculate the economic cost of this time to be over $160B. Tragically, it does not have to be this way. Even at current technology readiness levels, the Rand Corporation estimates that autonomous vehicle adoption could save thousands of lives over the next decade. In addition, transportation experts project that AVs will unlock hundreds of billions of dollars, mostly in the form of reduced collision costs and the reclamation of productive time.

However, these massive societal benefits come with their own societal costs. According to US census data, driving a truck is the most common job in 29 states, and in total there are 3.8 million truck and taxi drivers at risk of losing their jobs. In fact, a Goldman Sachs study predicts that by 2030, AVs will be displacing up to 300,000 jobs per year.

Still, the largest threat to AV adoption will be public sentiment. A survey from the Wall Street Journal highlights that only a tiny sliver of the population (11%) today is willing to ride in a fully autonomous vehicle, despite them being substantially safer than the average human. Researchers from the Max Planck Institute suggest this is largely due to intrinsic psychological biases that convince us that we’re safer drivers than we actually are.

If Autonomous Vehicles are to be successfully integrated into our society, both the economic and safety concerns must be addressed proactively and effectively. Indeed, all members of the AI community, not just the AV community, should be laser focused on getting this right. If the technology fails to deliver the promised safety benefits, or if the war of public sentiment is lost, the consequences will spillover to all future applications of AI. Considering the mammoth humanitarian capacity of future AI technologies like personalized healthcare and synthetic biology, any unnecessary delay of progress should be seen as very costly.

Fortunately, the inverse is also possible. Strategic roll-out and public education can improve consumer trust. Bold social policies can mitigate job displacement. Successes can be publicized, and inevitable concerns can be addressed just as publicly. Autonomous Vehicles can break down mental barriers around AI and establish precedent for this new technology serving as a public good.