The United States has built a transportation system that, along with its many benefits, is also deadly, expensive, and damaging to the environment and public health, while also failing to serve much of the public much of the time.

The numbers are grim: more than 32,000 people are killed and 2.3 million are injured in motor vehicle crashes on our roads each year, at an economic and societal cost of more than $800 billion annually. Another 50,000 lives are cut short each year due to air pollution from motor vehicles. The dangers of global warming continue to mount, fueled in part by a U.S. transportation system that produces more greenhouse gases per capita than that of any other large industrialized country. And all that destruction and pain – much of it concentrated in our cities – has been in service of a system that leaves large parts of society without access to convenient mobility and consigns millions of others to spend long periods of their lives stuck in traffic, at incalculable cost to their physical and mental health.

Autonomous vehicles have the potential to address many of those problems. But will they?

Before we can answer that question, we first need to confront an uncomfortable truth: Many of the problems we hope will be solved by autonomous vehicles can already be solved, or at least ameliorated, without them. Right now.

Designing streets for slower speeds and providing better facilities for pedestrians and cyclists can already reduce traffic deaths and injuries. Air pollution and greenhouse gas emissions can already be reduced by providing safe, convenient alternatives to personal driving, requiring cleaner cars, and building communities where low- and zero-carbon modes of transportation are viable options. Congestion can already be curbed through smart pricing of roads and parking.

Why are we not already addressing these problems with every tool in our toolbox? Why is it that, when policy-makers are asked to choose between a bike lane and a parking space, a traffic calming measure and a few seconds of motorist delay, or requiring drivers to pay the full cost of their use of the transportation system versus spreading those costs broadly across society, they so often choose to prioritize the use of personal cars over other societal needs?

And what makes us think that the advent of new technology, in and of itself, will lead us to make different choices in the future?

One answer might be that driverless cars will simply cause the messy trade-offs at the heart of our transportation debates to melt away. Technology will allow us to have all the rapid personal mobility we want and more, with none of the nasty, costly side effects. We can have our cake and eat it, too.

There may be some validity to this when it comes to the potential of driverless cars to protect the safety of vehicle occupants – it is hard to envision a future system that could be any less safe than what we have now. But history gives us many examples of technological advances that could have delivered great social benefits, but did not.

From the 1980s to the 2000s, for example, automakers made continual improvements to internal combustion engine vehicles sufficient to improve the fuel economy of those vehicles by, according to one study, as much as 60 percent. In reality, however, vehicle fuel economy barely improved at all. The reason: automakers used those technological advances to build bigger, heavier and more powerful cars, while using their political clout to oppose stronger fuel economy standards that might have translated those technological improvements into reductions in oil dependence and greenhouse gas emissions.

There is little reason to believe that, absent strong and coherent public policy guidance, the transition to autonomous vehicles will take a more enlightened or socially conscious path than previous automotive transitions. After all, automakers continue regularly to market cars based on the vehicles’ ability to do things (such as travel 205 miles per hour) that would be reckless and plainly illegal on any American public road.

Building a transportation system that is safe, sustainable and efficient requires more than technological progress; it also requires a policy architecture and transportation culture that prioritizes those goals, whether the vehicles traveling within that system are operated by humans or not.

Building that architecture now, even in advance of the commercialization of autonomous vehicles, can ensure that when they arrive – and in whatever form they arrive – they are born into a transportation system ready to maximize their benefits for everyone.

If we want an autonomous vehicle future that is safe for everyone, the best way to achieve it is to enshrine safety as a paramount concern of transportation policy now. If we want a future system that meets our goals for climate action, the best way to do so may be to create frameworks and support technologies and strategies that move us toward decarbonization of the transportation system today. And if we want to ensure that a future of autonomous vehicles will not be one of more congestion and more sprawl, we might adopt regulatory and pricing reforms that encourage efficient use of the transportation system and smart land use practices before that future is suddenly upon us.

The advent of autonomous vehicles provides a once-in-a-lifetime opportunity to renegotiate the bargain between American society and the car, as well as the potential for dramatic and positive transformation of our transportation system. We cannot afford, however, to wait until autonomous vehicles rule the roads before we set the rules of the road. The time to start laying the policy groundwork for the transportation system of the future is now.