What Autonomy Can Deliver
Once cars can drive themselves, do we ever need to get in them again?
If you ask the Internet for an image search of the future of mobility, it will return a lot of slick, futuristic cars that are, inevitably, fully autonomous. This has been the dominant narrative for the past few years and is still high on the Valley hype cycle. But the vision of replacing the driver with a crown of sensors and a supercomputer in the trunk is proving both challenging and not that inspiring.
The real prize of the autonomous economy is not a self-driving car, it’s the freedom humans will enjoy with not needing a car at all. The car is usually a means, not an end. In that light, making a better car is solving the wrong problem.
One real value of autonomy is freedom from the burden of schlepping our belongings. Specifically, carrying all those belongings with us, even when we don’t need them in that moment. We’ve grown accustomed to carrying multiple bags, with options for multiple contingencies. Chargers, batteries, computers, phones, make-up, medicine, spare clothes, extra shoes, snacks, water, coffee. As we have become more mobile, we continue to find ways to extend our time out of the house. The cycle is self-perpetuating. And every thing we carry needs a means to carry it.
The list of things that need to be moved grows relentlessly. From getting to work to getting kids to school to getting groceries, dry cleaning, picking up the dog, the one commonality is that we are surrounded by people and things that need to move from where they are to where they need to be. And spending time doing that moving feels like drudgery because most of it is. Part of what makes it feel so burdensome is that we run those errands and chores with our cars, which are often not the best tool for the job, just the one we have at our disposal.
This begins to describe why the autonomous car has captured our collective imagination. Imagine not having to drive to pick up the dry cleaning? Imagine if there was a service that could safely take our kids from school to practice? Immediately the image of those services conjures a car because that’s what we use today. Why deploy a 4,000–6,000-pound metal box that is capable of going 100+ mph and surviving a 35 mph frontal crash to drive three miles and pick up 10 pounds of groceries? The amount of energy used is vastly in favor of moving the car, not the groceries.
Multiply this inefficiency times every errand, commute, and chore and you end up with pollution, congestion, and a catastrophic waste of resources.
We see the same inefficiency when it comes to the space that cars use. We cede most of our urban public space for the storage and moving of cars and force the people who live in cities to live in the margins. We make enormous sacrifices in order that we have access to the “freedom” that cars permit. But that freedom, increasingly, is more of a trap.
Since most people have just one car, that car is purchased for an occasional extreme use case, not the median use case. For example, if you are going to drive in snow, you buy four- or all-wheel drive. If you camp once or twice a year, you make sure you have space for the whole family and all the gear. Having the ability to move all your stuff and your friends/family drives that decision. It’s a rational choice but it’s also the place where autonomy can really change the way we make decisions for the way we live because autonomy can solve for the edges and allow us to make a choice that is right for the median.
Imagine what happens to our world when we can move independently from our belongings. What if our mountain bike could meet us at the trailhead, or if our stroller could meet us at the airport? What if we only traveled to the places we wanted to go, and could cut out the places we didn’t?
Two things happen once we have the ability to move independent of our things: we expand the ways in which we can move and we dramatically reduce the energy required to move. We can right-size every trip so instead of being prepared for the extremes and paying the price with every subsequent journey, we only use the minimum amount of mobility we need.
If my yoga bag can just meet me at the studio after work and my dry cleaning will arrive at my house when I get home, then my options for getting to work expand dramatically. If I only have the things I need with me, then a bike might be a much more viable option. Or an electric skateboard. I might just walk.
Meanwhile, my yoga bag actually has eight hours to get where it’s going. It doesn’t need a ride in a private car. It could easily make it by delivery bot that can trundle along slowly and safely (maybe with other yoga bags) and it would use a fraction of the energy.
Autonomy also allows us to dramatically increase the time window in which things happen. Most traffic is a result of everyone trying to arrive at work at the same time. Most roads are sufficient for the number of cars that there are, but not if all of the cars go at the same time. So errands that aren’t bound by time can happen overnight, or be optimized for the right time. We needn’t constrain ourselves to roads, even as the sky is starting to be a viable medium for certain errands. If I’m even slightly organized, the kit that I need for Wednesday morning yoga could start making its way to the studio on Tuesday night.
We use an astonishing amount of energy just moving things from kids to yoga mats, to the cars themselves. Autonomy frees us from the dual constraints of time and people required to do that moving and once that is gone, the entire calculus changes. Traffic diminishes. Pollution is reduced. Congestion is a thing of the past. Traffic deaths decrease. All of the promises of autonomy still hold true.
Moving things isn’t anywhere near as sexy as the promise of having cars that drive themselves, but it’s a hell of a lot more useful. It’s also an easier problem to solve. Because of the Trolly Problem and the expectation that autonomous cars move at the same speed as our current ones, we still have a lot (a lot) of trust and safety issues to resolve before self-driving cars make a measurable impact on our lives. But a small, slow, autonomous box that is quietly moving yoga mats in the middle of the night doesn’t have any of the same safety concerns.
This also serves the purpose of giving the engineers who are trying to solve “the mother of all AI problems” a safer and more relevant sandbox to work in. They could deploy hundreds of actual bots on actual streets that are creating real market value without asking people to get inside them.
The promise of freedom in an autonomous world is still real and at least as valuable as it’s been promised. If it seems like we’re not getting any closer to that goal, it might just be that we’re not looking at the right finish line.