The road ahead for self-driving vehicles is littered with complications

Recent developments in San Francisco highlight some of the controversies over self-driving cars.

The California Public Utilities Commission authorized a 24/7 test deployment of 200 self-driving taxis. After an accident, the Motor Vehicle Department cut the test in half.

It’s worth taking a closer look.

Few new, tech-driven products have the potential to affect our lives as much as fully autonomous vehicles (FAVs). But history is also littered with amazing new, tech-driven products that were either short-lived or failed. Passenger blimps, cassette tapes, and highly-touted PDAs like Apple’s Newton are examples of dead-end innovations that either flopped or were quickly replaced by something better.

As with many society-changing innovations, America’s FAV controversies are multi-level, complicated, disputed and driven by self-interest. The most basic controversy is whether FAVs actually work. The question is important, as human lives potentially hang in the balance.

This question has many parts. Fundamentally, FAVs are the merger of (mostly electric) automobiles, advanced sensors, closed radio communications systems and artificial intelligence. FAV advocates explain that each of these technologies is reliable and has been operating and tested for at least a decade. Critics contend that each technology element has flaws, and that it only takes one of them — or the seamless integration of all of them — for the entire FAV to fail.

This debate is further complicated by the notion “don’t compare me with the perfect; compare me with the alternative,” which in this case means FAVs may not be perfect but they are safer than humans driving cars. So, advocates of FAV reliability often lead with the argument that, compared with human drivers, FAVs will save lives.

Unlike most tech-driven innovations, the replacement of human drivers by AI-driving FAVs raises significant employment questions. The Federal Transportation and Labor Departments estimated in 2021 that 4.6 million Americans “earn their livelihoods in occupations where driving a car, truck, or other vehicle is the primary job responsibility.” And there are over 1 million Uber and Lyft drivers in the U.S., many of whom are part-time. Regardless of methodology, it’s difficult to avoid the conclusion that millions of U.S. drivers’ jobs could someday be partially or entirely replaced by FAVs.

Because FAVs don’t sleep, take vacations, take coffee breaks, require health insurance, get medically sick or require withholding taxes or human resource departments, the conversion from human to AI drivers offers enormous potential to increase the productivity and efficiency of transportation. This not only offers the economy and existing transportation-centered businesses benefits, but it opens the door to entirely new FAV businesses.

Not surprisingly, many of the 5 million or so Americans who primarily drive for a living don’t necessarily see it that way, and so some of them and their unions have been actively opposing federal and state policies that would encourage FAVs. Underlying these economics is the potential for FAVs to replace conventional personal-use automobiles, many of which today have already been partially automated, such as my own new hybrid. McKinsey recently estimated that the potential U.S. market for personal-use FAVs could be $300 to $400 billion by the mid-2030s, a large new market for manufacturers.

The tension between drivers/unions and the FAV industry has been on full display recently in trend-setting California, where the state Assembly almost unanimously approved a law that requires human truck drivers in long-haul trucks. Although the final outcome in California remains uncertain, this battle starkly reveals the employment controversy that underlies the FAV debate.

The controversies in California over FAVs also reveal the pure governmental complexity of regulating FAVs. For many decades, America’s legal structure regulated automobiles, trucks and buses on the foundational notion that there is a licensed vehicle and a licensed human driver. To oversimplify, vehicle safety and fuel economy are regulated by federal agencies; liabilities, driver and vehicle insurance, licensing vehicles and drivers and some fuel efficiency and vehicle safety standards are regulated at the state level; and many road regulations are set by local governments. Trying to fit driverless vehicles into this matrix is quite difficult, mostly because essentially every regulation currently presumes there is a human driver.

Without coordination between the levels of government, disjointed FAV regulation is likely. As a result, some have suggested that the federal government should find a way to effectively pre-empt state and local governments in order to introduce a nationwide legal framework for FAVs. Even this approach, however, requires complicated coordination between many agencies within the Departments of Commerce, Transportation and Labor and their respective congressional oversight committees.

Finally, international geopolitical controversies play a role. Many advocates of government encouragement for FAVs base their advocacy in part on the need for America to lead this emerging global technology and market in our contest with China. For economic, military, political and soft power reasons, their view is that if the U.S. doesn’t lead and dominate the FAV market, China will.

Thus, any of the risks resulting from support for FAVs must be considered in the context of the military, economic and political risk of Chinese domination. Not surprisingly, others do not see the global picture as zero-sum and emphasize the welfare of average Americans as their paramount consideration.

After considering federal FAV legislation for five years, Congress appears hesitant and divided. The overall policy road ahead seems complicated and less than clear.

Roger Cochetti has served as a senior executive with COMSAT, IBM, VeriSign and CompTIA. A former U.S. government official, he has helped found a number of nonprofits in the tech sector and is the author of textbooks on the history of satellite communications.