Overview of Interesting Times: Why Are We Still Driving?
This episode of Interesting Times features Ross Douthat in conversation with transportation writer Andrew Miller about the rise of self-driving cars and what a more automated driving future could mean for safety, mobility, economics, culture, and personal freedom. Miller is broadly optimistic about robotaxis and partial autonomy, arguing that the technology could save lives, free up time, and make transportation cheaper and more efficient. Douthat, meanwhile, presses on what might be lost if driving becomes less central to American life.
Main Topics Discussed
Why self-driving cars matter
Miller’s core argument is that autonomous vehicles could deliver two huge benefits:
- Fewer deaths and injuries: He cites roughly 40,000 road deaths a year in the U.S., most caused by human error.
- More free time and attention: If people no longer have to actively drive, they regain hundreds of millions of hours for work, leisure, or anything else.
The timeline for adoption
Miller predicts that by around 2035, self-driving taxis could be normal in many North American cities, especially major metros. He expects:
- Small autonomous fleets to expand city by city
- Robotaxis to become as routine as Uber is now
- Adoption to happen first in dense urban areas, then potentially in suburbs
Waymo vs. Tesla
A major part of the conversation compares the leading self-driving approaches:
- Waymo uses multiple sensor types, including cameras, radar, and LiDAR, which makes its cars more reliable in difficult conditions but also more expensive.
- Tesla is betting on a cheaper, camera-heavy approach that could scale faster if it works, but Miller sees it as less proven and potentially less safe.
The tension is essentially:
- Waymo = safer, but costlier
- Tesla = cheaper, but riskier
Liability is the big bottleneck
Miller argues that liability is the most important unresolved issue.
- If self-driving systems are controlling the car, manufacturers should accept responsibility when the system is at fault.
- He believes this is the clearest way to move the industry forward.
- Companies like Waymo have accepted this to some degree; Tesla has been more reluctant.
He also notes that regulators are holding autonomous vehicles to a higher standard than human drivers, which may slow adoption but is understandable given public distrust.
Remote assistance is not the same as full autonomy
The episode digs into a subtle but important distinction:
- Some autonomous systems still rely on human remote assistance when they get confused.
- Miller argues that this is not truly “driving,” but rather human support for machine decision-making.
- Passengers can usually contact a human operator if something goes wrong, but they generally cannot directly override the vehicle in real time.
Safety concerns and weird accidents
The conversation acknowledges that self-driving accidents can feel especially unsettling because they are often unusual or “inhuman” in their behavior.
Examples discussed include:
- A Cruise incident that badly injured a pedestrian and helped kill the company
- Waymo incidents involving a bodega cat and a child being hit
- Tesla crashes involving white tractor-trailers that were difficult for cameras to detect
Miller’s view is that these problems should improve as systems get more data, but they also justify cautious rollout.
Broader Social and Political Implications
Public transit could be helped—or harmed
Miller offers two possible futures:
Good scenario
- Robotaxis become cheap and widely available
- Households need fewer cars
- Parking demand drops
- Roads are safer
- People reclaim time and money
- Electric autonomous fleets help reduce emissions
Bad scenario
- Robotaxis increase congestion without replacing enough private car use
- Public transit loses riders and enters a death spiral
- Lower-income riders are left with worse mobility options
A culture war issue in the making
The episode highlights that self-driving cars could become politically divisive.
- Blue states may be more skeptical because of labor, privacy, and regulation concerns
- Red states may be more open to the technology in the name of consumer freedom and innovation
- But the irony is that the most libertarian states may end up enabling the technology that reduces human driving most aggressively
Privacy and security concerns
Because self-driving cars are constantly collecting data, they raise new worries:
- Who controls the data?
- Could companies or governments misuse it?
- Could hackers exploit autonomous fleets?
Miller argues that while these threats are real, they are probably less dangerous than softer, more obvious targets like utilities or power grids.
What Driving Represents, and What Might Be Lost
Douthat pushes on the cultural meaning of driving:
- It is tied to American freedom, independence, and adulthood
- Learning to drive has long been a kind of rite of passage
- Driving is also a form of embodied knowledge: using your body, reflexes, and attention in the real world
Miller doesn’t deny that something is lost, but he argues that:
- We shouldn’t ask driving to preserve all forms of human meaning
- If a technology can save lives and restore time, society should adopt it
- The deeper problem of disembodiment in modern life should be solved elsewhere, not by preserving dangerous driving as a cultural anchor
Closing Thought
The episode ends on a pragmatic, slightly ironic note: the future of self-driving cars may be determined less by whether the technology works than by whether society is willing to accept it, regulate it, and adapt to it. Miller sees a future where autonomous driving becomes normal, safe, and mostly boring—and where the real challenge is not engineering, but politics, liability, and cultural acceptance.