Carmakers fret their industry is on the brink of huge disruption

I don't have the data, but that's intuitively correct. Time does matter some when it comes to the corrosion of the underframe (in states that salt roads), dried gaskets, weather damage to paint. But miles matter for everything else including body damage from accidents.

> I've always been told that mileage--not time--is the main variable.

And my experience is exactly the opposite. I have never had a car die from mileage. I've had several die from time.

In my experience cars last about 10 year, 15 if you are careful. How many miles you drive during that time is mostly irrelevant.

Personally I think cars should track hours when turned on, not miles.

A lot of cars to destroyed in car accidents every year. If you assume ~50,000 fewer totaled cars per year that's a lot of extra miles.

Also, several cars have made 1+Million miles as I the odometers roles over. Considering many people drive less than 8k miles per year that's a lot of driving.

PS: The real issue is both age and driving reduce value. But, age is automatic as soon as you know the cars model year so people care about high vs low mileage because they are already buying an old car.

If the car is going to be used a lot, then it becomes economical to use more durable parts. For example components could be used that last four times longer while only costing twice as much. This is pretty much economies of scale, and it will be far easier for a company like Uber to get them, than individuals.

It entirely depends on what kind of mileage and what kind of time. 100tkm in city traffic only could easily be more devastating than 500tkm on highway. (Or some similar-ish proportion.) Conversely, time spent in controlled storage, with appropriate precautions taken before and after, doesn't necessarily count much. Yet time spent in active use even with low mileage (say, a single one-stretch 100 km trip per week) will wear out parts and components. Changes in temperature and humidity have an effect there. I've also observed common replacement cycles mostly regardless to mileage for some parts: a ten-year car is likely to need similar parts replaced regardless of whether it has 100tkm or 250tkm on it. Some parts are more worn by miles, some parts more by time in active use.

"we really cannot have 7 billion people drive cars" ... I think we will have to... for a while.

Education is the best way of reducing population. Cars (or any type of personal transport) is crucial, moving the effort from humans to machines, leaving the human with more time and options. This leads to education and caring about the environment, politics etc.

So for a while we'll have to have 7 billion people using cars (or at least that's what the goal should be) then this will drop to 6 billion, then 5 then 4 etc.

We can power cars using solar. Terrestrial or orbital.

I see this exact sentiment all the time. Even people who have a favorite brand will gripe angrily about getting ripped off on the purchase and how repairs are so needlessly expensive.

So true. I have zero loyalty to my car maker. Or, to formulate it positively, in all the years of owning a car I have only once been positively surprised about the service I got.

Self-driving cars drastically reduce the total number of cars necessary, and, as the article notes, limit the ability of manufacturers to upsell features. If the fleet of cars on the road were largely self-driving, there is no way it would support the present variety of car manufacturers. I would hazard that most would go bankrupt.

Seconding shakethemonkey's comments and would ask: how much thought do people put into the car they buy versus the cars they rent?

The car companies have enjoyed brand loyalty and the ability to sell feature bundles because many people associate their car with their identity. When you say “Mustang owner”, “Subaru driver”, etc. specific images come to mind, and across all cars you see things like people paying considerably more to e.g. have a fuel-inefficient engine because there are a few times where they can step on the gas and enjoy the rapid acceleration. Manufacturers made billions selling SUVs to people who thought safer, cheaper minivans were too boring.

That doesn't completely go away with self-driving cars – someone with 4 kids will still have different needs than the avid bicyclist or single urbanite — but a lot of it will become less important if people start thinking about their car as, say, the place where they play games or catch up on Netflix while they head home from work. Why pay Ford for the premium sport package when you can put the money into a faster mobile data plan or a better tablet? They certainly don't go out of business in that scenario but the business becomes tighter and less exciting if it becomes more of a commodity where people generally just buy whatever's cheaper.