Driverless cars and auto insurance

[NormR]

Any thoughts on the impact of driverless cars on the auto insurance business? 

[rkbabang]

Short term not much, longer term 10-30 years, I think the entire industry will change significantly.  I don't think the economics will favor owning your own car for 98+% of people.  A car is a huge expense and spends the vast majority of its life sitting in a driveway, garage or parking lot while it depreciates.  If you could quickly and easily get a car to pick you up (in minutes) and only pay for the time you actually spend inside one that just seems to make more economic sense.  Therefore what insurance companies will be insuring is the corporate fleets of the large car sharing companies.

[jay21]

Short term not much, longer term 10-30 years, I think the entire industry will change significantly.  I don't think the economics will favor owning your own car for 98+% of people.  A car is a huge expense and spends the vast majority of its life sitting in a driveway, garage or parking lot while it depreciates.  If you could quickly and easily get a car to pick you up (in minutes) and only pay for the time you actually spend inside one that just seems to make more economic sense.  Therefore what insurance companies will be insuring is the corporate fleets of the large car sharing companies.

 

The future of the auto insurance industry is more a function of expected auto loss than anything else imo. The biggest threat is driverless cars are better drivers than humans (great social outcome tho).

[vinod1]

This would be too difficult to predict as there are going to be lot of 2nd order and 3rd order effects.

 

When we have a true driverless car, it might actually increase car usage very significantly

 

- Segments of populations that cannot drive a car (handicapped, older citizens, cannot get a driver's license, etc) would now be able to use them.

- People would be more inclined to use them to drive long distances commuting to work, etc as they can be working in the car

- Cars could take some of the market share away from public transport, airlines, etc as it would be more convienent

 

The impact would also depend upon how the regulation would evolve and it might favor one industry over another.

 

VInod

[Jurgis]

This would be too difficult to predict as there are going to be lot of 2nd order and 3rd order effects.

 

+1 agreed.

[rkbabang]

This would be too difficult to predict as there are going to be lot of 2nd order and 3rd order effects.

 

When we have a true driverless car, it might actually increase car usage very significantly

 

- Segments of populations that cannot drive a car (handicapped, older citizens, cannot get a driver's license, etc) would now be able to use them.

- People would be more inclined to use them to drive long distances commuting to work, etc as they can be working in the car

- Cars could take some of the market share away from public transport, airlines, etc as it would be more convienent

 

The impact would also depend upon how the regulation would evolve and it might favor one industry over another.

 

VInod

 

Of course, but it is still possible to make some reasonable predictions. I think car usage would increase dramatically, but the number of insured cars would be much less than now.  You wouldn't need 2-4 cars per family anymore, a transportation company could serve thousands or tens of thousands of people with hundreds of cars.

 

Also while people will always prefer riding in cars rather than trains/buses, there is still the fact that if everyone wants to get to work by 9AM the roads will be congested in the morning.  Autonomous cars can ride closer together and do not rubber neck accidents on the other side of the highway, or freak out because they saw a break light in front of them, but there is still a limit to how much traffic a road can handle before things slow down.  And all it takes is one or two human drivers on the road to screw everything up.  I wonder if human drivers will even be allowed?  I don't think we will be completely rid of traffic (which means some people will still use public transportation such as commuter rail) until we have autonomous flying people carriers.  There is plenty of room in the air to carry everyone where they want to go if the machines are smart enough not to crash into each other. 

 

Also there might still be room for long distance bus/train transportation for cost reasons.  I'm sure it will be much cheaper to take a bus from Boston to NYC than a car.  Especially if the car company doesn't operate in both locations.

 

Things this will kill is roadside gas stations, airport transportation companies, taxis, intra-city buses/subways.

 

[Guest Schwab711]

As of now, I don't think there will be a noticeable effect on the auto insurance industry in my lifetime (or at least 50+ years, to use less hyperbole). I don't think the incentives for conversion are present for the majority of Americans.

 

I don't think driverless vehicles will see consumer adoption at meaningful rates for a significant period of time (generations). I think adoption (if it happens at all) will be very slow, with a lot of volatility in acceptance and a ton of failures along the way. Any early software/sensor failure could possible halt progress for years.

 

Just to get the manufacturing capacity required for mass-adoption in the US will require multiple decades of investment in new/refurb production plants, presumably by the current major manufacturers who have legacy interests that they are still trying to generate returns from (to avoid major writedowns). That doesn't even touch the capacity concerns for sensors and batteries that will be necessary (assuming these driverless cars are at least partially EV). The country can't even manufacture EV vehicles at a 10% market share. Driverless cars will likely be many years down the road due to investor incentives.

 

There are also tremendous technological hurdles that need to be resolved for adoption rates to increase from 0.1% to 1%, much less 10% or ~100%. Regulations ensuring safety will likely slow adoption as much as capacity constraints. The technology is not necessarily easy either. Driverless cars will have to progress from being able to handle the less-predictable human decisions and movements of other drivers to a hybrid system that can handle human drivers while simultaneously communicating with other driverless vehicles. Their communication system will become exponentially more complex as adoption increases, which is a tough problem economically. Privacy concerns will also slow progress as other drivers will need to consent to their location being used for driverless technology to work.

 

Finally, I'm not sure the economics make sense for the majority of people. Although driving costs time, it's not like people are clarmoring to take the bus even if it means they can work while they travel, save money, and help the environment. Also, I highly doubt regulations will allow folks to be able to look away from the road for extended periods in live environments. What if the software or connectivity fails and manual driving is required, but the passenger does not assume control of the vehicle?

 

I am sure my predictions are going to be absolutely wrong, but I just don't see how wide-spread adoption is possible. I think vehicles that don't use gasoline (or high-mileage vehicles) and significant investment/innovations in public transportation are much more realistic transportation changes in my lifetime. I think driverless cars have a better chance of adoption as an extension of public transportation where you request a car to transport you to your destination from your door. No one would own these vehicles and you'd pay-per-use.

 

---

 

As a side note, I wonder if the majority of board members who do see wide-spread adoption in the near-future are from major cities onthe west coast. The economics certainly make more sense out there due to the general differences in public transportation systems.

[Jurgis]

Schwab711, I completely disagree with you.

 

We will have ~100% penetration of no-human-in-loop cars in 30 years or less.

 

As an aside we will have some form of human immortality in 40 years or less.

 

Also superhuman level AI in similar time frame.

 

(Unless we have a civilization extinction even in the meantime).

 

The future's so bright we gotta wear shades.  8)

[EliG]

Bank of England blog:

 

Driverless Cars: Insurers Cannot be Asleep at the Wheel

 

 

[rkbabang]

Schwab711, I completely disagree with you.

 

We will have ~100% penetration of no-human-in-loop cars in 30 years or less.

 

As an aside we will have some form of human immortality in 40 years or less.

 

Also superhuman level AI in similar time frame.

 

(Unless we have a civilization extinction even in the meantime).

 

The future's so bright we gotta wear shades.  8)

 

I agree with you, I'm just not willing to put a guess on the time frame.  I hope you are correct.

[NormR]

Personally, I'd snap up a self-driving car that's fairly affordable and safer that a human-driven one.  Unlike most people, I count myself as a below average driver.  :) 

[Jurgis]

Personally, I'd snap up a self-driving car that's fairly affordable and safer that a human-driven one.  Unlike most people, I count myself as a below average driver.  :)

 

I consider myself above average driver (haha, lol), but I'd buy one in a heartbeat too. I don't see a point of me spending time, effort and mental energy driving. I'm always happy when someone offers to drive instead of me driving.

 

And even if I'm above average driver, the self-driving car would still be safer than me driving. :)

[Guest Schwab711]

Schwab711, I completely disagree with you.

 

We will have ~100% penetration of no-human-in-loop cars in 30 years or less.

 

As an aside we will have some form of human immortality in 40 years or less.

 

Also superhuman level AI in similar time frame.

 

(Unless we have a civilization extinction even in the meantime).

 

The future's so bright we gotta wear shades.  8)

 

I definitely agree on the candela level in the future :). I just think this specific solution is going to be tough to implement because it is not solely determined by the ability of companies to produce the car. I can't even imagine our society realistically removing fueling stations anytime soon. I really think we are moving towards a hybrid mix of EV and cars as we know them and/or cars as we know them are going to shift towards ultra fuel-efficient designs (which already exist - a legislative push would certainly revive them if there were ever the support). In 1969, it was a generally accepted notion that we'd colonize Mars by 2000. I just don't think it's possible to provide the safety folks are envisioning in such a short period of time. Just think of all those future connectivity accidents!

[Uccmal]

As of now, I don't think there will be a noticeable effect on the auto insurance industry in my lifetime (or at least 50+ years, to use less hyperbole). I don't think the incentives for conversion are present for the majority of Americans.

 

I don't think driverless vehicles will see consumer adoption at meaningful rates for a significant period of time (generations). I think adoption (if it happens at all) will be very slow, with a lot of volatility in acceptance and a ton of failures along the way. Any early software/sensor failure could possible halt progress for years.

 

Just to get the manufacturing capacity required for mass-adoption in the US will require multiple decades of investment in new/refurb production plants, presumably by the current major manufacturers who have legacy interests that they are still trying to generate returns from (to avoid major writedowns). That doesn't even touch the capacity concerns for sensors and batteries that will be necessary (assuming these driverless cars are at least partially EV). The country can't even manufacture EV vehicles at a 10% market share. Driverless cars will likely be many years down the road due to investor incentives.

 

There are also tremendous technological hurdles that need to be resolved for adoption rates to increase from 0.1% to 1%, much less 10% or ~100%. Regulations ensuring safety will likely slow adoption as much as capacity constraints. The technology is not necessarily easy either. Driverless cars will have to progress from being able to handle the less-predictable human decisions and movements of other drivers to a hybrid system that can handle human drivers while simultaneously communicating with other driverless vehicles. Their communication system will become exponentially more complex as adoption increases, which is a tough problem economically. Privacy concerns will also slow progress as other drivers will need to consent to their location being used for driverless technology to work.

 

Finally, I'm not sure the economics make sense for the majority of people. Although driving costs time, it's not like people are clarmoring to take the bus even if it means they can work while they travel, save money, and help the environment. Also, I highly doubt regulations will allow folks to be able to look away from the road for extended periods in live environments. What if the software or connectivity fails and manual driving is required, but the passenger does not assume control of the vehicle?

 

I am sure my predictions are going to be absolutely wrong, but I just don't see how wide-spread adoption is possible. I think vehicles that don't use gasoline (or high-mileage vehicles) and significant investment/innovations in public transportation are much more realistic transportation changes in my lifetime. I think driverless cars have a better chance of adoption as an extension of public transportation where you request a car to transport you to your destination from your door. No one would own these vehicles and you'd pay-per-use.

 

---

 

As a side note, I wonder if the majority of board members who do see wide-spread adoption in the near-future are from major cities onthe west coast. The economics certainly make more sense out there due to the general differences in public transportation systems.

 

Schwab, I agree with most of this. 

 

Folks who work in technology which are the majority on the board vastly overestimate the difficulty of implementation, and the reliability of a driverless car system. 

 

Before I ever use a driverless car I want to see an entire fleet operate 90 days in Rome Italy, and Kathmandu, Nepal without a single collision of any kind.  Since I have yet to meet a computer system that doesn't crash, or a connectivity system that doesn't get blocked from time to time I think we are looking out at least 50 years. 

 

Call me a skeptic, but I have yet to see an infallible technology of any kind, or even nearly infallible.  A handful,of ugly incidents will put this idea on ice for decades. 

 

And we know how companies and governments and people like to cut corners.  There is no reason to suggest this wont be done in the driverless car business as well.

 

Finally, the world never unfolds as we expect.  Nuclear Fusion anyone. 

[redhots]

Washington Post:  Here’s how much a self-driving car could save you on car insurance

http://www.washingtonpost.com/blogs/innovations/wp/2015/07/17/heres-how-much-a-self-driving-car-could-save-you-on-car-insurance/?tid=sm_tw

 

Most of the cost savings from self-driving cars come from the expected near elimination of accidental collisions (we used a 90 percent reduction figure).

[Travis Wiedower]

I don't know why people demand driverless cars to be perfect. There will of course be glitches and issues but look at the driving record of humans. If Tesla has a legit driverless car in 10 years I'm confident it'll be significantly better than humans, but I don't expect them ever to be perfect. Of course one company producing a good driverless car to the entire economy switching over is another 10-30 years probably.

 

I also think luxury (especially super cars like Lamborghinis) will decline in popularity. If driverless cars become ubiquitous and human drivers are outlawed, what's the point in having a Lamborghini? It still looks bad ass of course but the enjoyment of driving it is no longer there. Maybe there will still be race tracks and such you can take them to but that'll be a very small minority of (rich) people who are that into it.

 

I don't know, predicting these things is incredibly difficult. Sure hope it happens over the next few decades though as I think watching the transformation will be fascinating.

[ERICOPOLY]

I predict more alcohol will be consumed when driverless cars eliminate the "designated driver".

 

Also, currently I see people have one drink at the bar (or at lunch) before driving home.  Why not have three instead?

 

Driverless cars are therefore likely to drive sales of booze and healthcare.

 

So if you own a pub, you'll probably see your margins expand as you can sell more drinks per customer.

 

I bet this is already happening with Uber and Lift -- would love to see the data.  I drank an extra few last night because we used Uber to drive only a mile across the neighborhood to a dinner party.  There were more liquor sales generated last night and it's 100% thanks to Uber.

[Uccmal]

Now we know why Eric stopped managing some

of his money, anyway.  Its the ole booze problem rearing its ugly head.

 

This article is interesting:

 

http://paleofuture.gizmodo.com/no-tech-adoption-is-not-speeding-up-1565326373

 

Some of the comments are really thoughtful as well.

 

I used a motorola phone in 1989 in a work truck.  I recall some Tv show from the mid 70s where someone had car phones - Charlie's Angles or Vegas (robert Urich) maybe. 

 

Driverless Vehicles - 30-50 years to mass adoption, if ever.  Things change, as in, the entire idea may get supplanted with other ideas such as the Hyperloop or some equivalent of people moving.

 

And, Jurgis, I aint gonna be downloaded to any computer where some a-hole will create dozens of duplicates of me, and screw up my code.  Id rather be dead than suffer that fate. 

 

[Jurgis]

Call me a skeptic, but I have yet to see an infallible technology of any kind, or even nearly infallible.  A handful,of ugly incidents will put this idea on ice for decades. 

 

We are back to the strawman argument.  ::)

 

Yes, everybody stopped driving GM cars "for decades" because of a "handful,of ugly incidents" with their ignition systems.  ::)

 

And obviously nobody buys cars with cruise control and antilock brakes, because the technology is not "infallible".

 

And, Jurgis, I aint gonna be downloaded to any computer where some a-hole will create dozens of duplicates of me, and screw up my code.  Id rather be dead than suffer that fate.

 

You prefer to die, that's your choice. Have fun.

[rkbabang]

"Driving in the Networked Age" by Reid Hoffman

[Pelagic]

I tend to think driverless cars are going to lead to less total cars. Where a family today might have two or three cars because of the need to leave them at their workplace or school while there, a single driverless car can act as a taxi for each family member shuttling them around town as needed. It will be interesting to see how auto insurers deal with driverless cars, I expect that if they become universal, there will be a large premium for individuals who want the option to drive the old fashioned way occasionally.

 

The fleet of driverless cars on call is an interesting idea but I think people will still want a personal car, especially families living outside the heart of major metropolitan areas. The real question is will your driverless car be able to hire itself out while you're working and don't need it, wonder what the insurers would think of that scenario.

[rkbabang]

I tend to think driverless cars are going to lead to less total cars. Where a family today might have two or three cars because of the need to leave them at their workplace or school while there, a single driverless car can act as a taxi for each family member shuttling them around town as needed. It will be interesting to see how auto insurers deal with driverless cars, I expect that if they become universal, there will be a large premium for individuals who want the option to drive the old fashioned way occasionally.

 

The fleet of driverless cars on call is an interesting idea but I think people will still want a personal car, especially families living outside the heart of major metropolitan areas. The real question is will your driverless car be able to hire itself out while you're working and don't need it, wonder what the insurers would think of that scenario.

 

There will certainly be people who own their own vehicle, but I think this will be at the wealthier end of the spectrum.  For most people it won't make sense.  Even for people who can afford it many will opt not to bother.  If you work 50 miles from home does it really make sense to have your car bring you to work then drive 50 miles back home empty to drive your kids to school, then drive 50 miles empty at the end of the day to pick you back up?  I think that as much as people above a certain age like driving and owning cars, it is more of a cultural thing that will not last very long (especially with younger people) if the economics doesn't make sense.  My son is 15 and I was talking with him about it and it'd be fine with him if driverless cars were available now and he never had to learn how to drive.  He'd rather watch youtube and let the car take him where he wants to go.  Sort of like now with me driving :)  Younger people simply don't have the same cultural attachment to cars and driving that people over 35 have.

 

[Dustin T]

I tend to think driverless cars are going to lead to less total cars. Where a family today might have two or three cars because of the need to leave them at their workplace or school while there, a single driverless car can act as a taxi for each family member shuttling them around town as needed. It will be interesting to see how auto insurers deal with driverless cars, I expect that if they become universal, there will be a large premium for individuals who want the option to drive the old fashioned way occasionally.

 

The fleet of driverless cars on call is an interesting idea but I think people will still want a personal car, especially families living outside the heart of major metropolitan areas. The real question is will your driverless car be able to hire itself out while you're working and don't need it, wonder what the insurers would think of that scenario.

 

There will certainly be people who own their own vehicle, but I think this will be at the wealthier end of the spectrum.  For most people it won't make sense.  Even for people who can afford it many will opt not to bother.  If you work 50 miles from home does it really make sense to have your car bring you to work then drive 50 miles back home empty to drive your kids to school, then drive 50 miles empty at the end of the day to pick you back up?  I think that as much as people above a certain age like driving and owning cars, it is more of a cultural thing that will not last very long (especially with younger people) if the economics doesn't make sense.  My son is 15 and I was talking with him about it and it'd be fine with him if driverless cars were available now and he never had to learn how to drive.  He'd rather watch youtube and let the car take him where he wants to go.  Sort of like now with me driving :)  Younger people simply don't have the same cultural attachment to cars and driving that people over 35 have.

 

I agree with you for the most part but I think there will be things that slow it down. First off public transportation just isn't as comfortable or clean as most people prefer their own vehicles to be. I think most of us would prefer to ride in our own cars over a typical taxi. People prefer their own space to communal spaces. Many a soccer mom will choose to transport their kids in a known clean vehicle than risk whatever disease the previous occupants of the week may have had.

 

Many people have hobbies or work requirements that a fleet of vehicles will be slower to provide. Hauling golf carts, four wheelers, kayaks, bicycles, home improvement supplies, and pets, and landscaping all to some degree require special equipment that will require specialized services. Certainly niche services will provide some of these, but there will always be atypical transportation needs.

 

We also enjoy the level of customization that comes with our own vehicles. Always having the seats adjusted just the way you like it, the presets on the radio, and other little creature comforts like heated seats. I also think working from a car would be much easier if it's customized to your liking. Not to mention entertainment I can imagine cars set up with wraparound IMAX type of entertainment systems or whatever else is used for entertainment in the future.

 

I think the tremendous savings will make penetration over 50% and will cause virtually every family to use them some. 

 

But I think the best correlation is probably traveling commercial vs. a private jet. Most of us can't afford a private jet, but the private car will be much more attainable for the middle class and possibly just as desired as the private jet is for the wealthy.

[rkbabang]

I tend to think driverless cars are going to lead to less total cars. Where a family today might have two or three cars because of the need to leave them at their workplace or school while there, a single driverless car can act as a taxi for each family member shuttling them around town as needed. It will be interesting to see how auto insurers deal with driverless cars, I expect that if they become universal, there will be a large premium for individuals who want the option to drive the old fashioned way occasionally.

 

The fleet of driverless cars on call is an interesting idea but I think people will still want a personal car, especially families living outside the heart of major metropolitan areas. The real question is will your driverless car be able to hire itself out while you're working and don't need it, wonder what the insurers would think of that scenario.

 

There will certainly be people who own their own vehicle, but I think this will be at the wealthier end of the spectrum.  For most people it won't make sense.  Even for people who can afford it many will opt not to bother.  If you work 50 miles from home does it really make sense to have your car bring you to work then drive 50 miles back home empty to drive your kids to school, then drive 50 miles empty at the end of the day to pick you back up?  I think that as much as people above a certain age like driving and owning cars, it is more of a cultural thing that will not last very long (especially with younger people) if the economics doesn't make sense.  My son is 15 and I was talking with him about it and it'd be fine with him if driverless cars were available now and he never had to learn how to drive.  He'd rather watch youtube and let the car take him where he wants to go.  Sort of like now with me driving :)  Younger people simply don't have the same cultural attachment to cars and driving that people over 35 have.

 

I agree with you for the most part but I think there will be things that slow it down. First off public transportation just isn't as comfortable or clean as most people prefer their own vehicles to be. I think most of us would prefer to ride in our own cars over a typical taxi. People prefer their own space to communal spaces. Many a soccer mom will choose to transport their kids in a known clean vehicle than risk whatever disease the previous occupants of the week may have had.

 

Many people have hobbies or work requirements that a fleet of vehicles will be slower to provide. Hauling golf carts, four wheelers, kayaks, bicycles, home improvement supplies, and pets, and landscaping all to some degree require special equipment that will require specialized services. Certainly niche services will provide some of these, but there will always be atypical transportation needs.

 

We also enjoy the level of customization that comes with our own vehicles. Always having the seats adjusted just the way you like it, the presets on the radio, and other little creature comforts like heated seats. I also think working from a car would be much easier if it's customized to your liking. Not to mention entertainment I can imagine cars set up with wraparound IMAX type of entertainment systems or whatever else is used for entertainment in the future.

 

I think the tremendous savings will make penetration over 50% and will cause virtually every family to use them some. 

 

But I think the best correlation is probably traveling commercial vs. a private jet. Most of us can't afford a private jet, but the private car will be much more attainable for the middle class and possibly just as desired as the private jet is for the wealthy.

 

I think you are correct it will always be better to own your own (I'd love a private jet) and for us who already own multiple vehicles, having only one will save us a ton of money already.  I'm just thinking about the average younger person who comes of age in the age of driverless cars.  As a college student then a young professional, then starting a family, buying a home, etc, there may always be something better to spend tens of thousands of dollars on than a private vehicle.  I tend to think private ownership will be an expensive luxury that many will aspire to, but fewer and fewer will be able to justify.  I think transportation companies will have different levels of service (think Uber black) that will guarantee nicer vehicles and state of the art entertainment.  Why buy your own which will be obsolete in 3 years, when the nicer car service companies always have the latest rides?  Have a car with a desk pick you up if you want to work, have a car with a wrap around screen pick you and the kids up for a long trip.  Have a small car bring you to the hardware store and have a truck pick you up from the hardware store (or better yet have a truck pick up what you ordered from the hardware store and meet you in your driveway without you ever leaving home). Same with the grocery store, you order the stuff you need online and have a vehicle go pick it up for you. 

 

It will certainly be interesting to see what actually happens.

 

[Jurgis]

There are couple issues with mass-taxi-no-ownership view of the future:

 

- Peak to trough demand/supply both in time and direction. Assuming most people still go to work at 9am and return at 5pm, will FutureUber maintain a huge fleet to satisfy demand at that time and direction? Compare this to midnight when nobody wants to take a car and most of the fleet is idle. Sure, FutureUber can apply surge-pricing, but that's not gonna make a regular-9-to-5 Joe happy, since they will have to pay surge pricing every day...

 

- Out-of-way places and destinations. If you want to get rental car in LA and drop it in Montana, you pay through the nose since the company does not need a car in Montana, they need it in LA. Similar situation will happen with people living or going to locations that are out of way. This also can be dealt with surge pricing perhaps, but I'm not sure that would make people happy either.

 

Also, I think there's a magical line of having car in 5 minutes or less. If you have to preorder a car and/or have waiting time over 10 minutes, the convenience drops quite a bit. Sure, in some cases it does not matter, but in some cases it's very annoying. Guaranteeing 5 minute car in peak demand or vice versa in out-of-way low demand location is going to be hard I think.

 

These might be solvable. :) I'm just thinking aloud.