The connected car may be the dumbest idea ever, but it’s not going away
arstechnica.com
arstechnica.com
Long ago their were PDAs before there were smartphones and they weren't always connected to the internet or capable of connecting to each other. Our smartphones are always connected, however, and are capable of so much more. Eventually people will think the same about cars. So features better than "Fitbit for cars" are coming.
- OTA updates. The reality of software: there will be bugs. With WiFi or LTE connectivity, you can fix fundamental issues by pushing updates to batches of cars in rolling cycles.
- e911. If you were to be in an accident, the vehicle can dial out to emergency services automatically with details about the vehicle's location and condition.
- Analytics. Every YC tech company most likely collects some form of data about how people use their product. Why? To be able to discover flaws and improve their offering. Car companies can build significantly better vehicles by collecting bits of anonymous telemetry from vehicles.
But, if we get rid of ownership of devices, we can have always available and always functioning devices that will arrive at your foot step in 5 minutes.
That said, I'm concerned about the avenue of control and monitoring this will provide, but it will happen either way, so we better deal with it properly instead of suing Uber in 3 years.
What this leaves out is people who genuinely enjoy driving as a pastime and would face a problem. I don't have a good answer for that except that private pilots also deal with it right now and we might be best served looking there for what did/didn't work.
I don't see any reason you'd need to have common ownership to implement 3rd party traffic control.
In reality, those corporations have complete control over the device. Only they can activate it for you, it's dongled to your account, you cannot sell it in aftermarket, they can deactivate it with a killswitch and remove or add functionality over air without your knowledge, they can spy on you and the device phones home without any way to completely deactivate it. It's very wrong.
People who want to own things, should be able to own things. We aren't slaves, right? People who want to rent/lease a service, may be able to do so as well.
> I don't see any reason you'd need to have common ownership to implement 3rd party traffic control.
You don't have to, but once vehicles work in a swarm and do not function stand-alone, you lose the benefits of pretend ownership. I suggested leasing a vehicle and thereby making it clear that you don't own it.
Also, except for circumventing any DMCA-covered "protection" on the firmware, you are not legally banned from modifying it. Use != copy. At best, you might be civilly liable for copyright violations.
The law that restricts what you can do with the AR-15 is (I believe) criminal law. Transmission power is similarly banned. (through executive regulation, I think?)
--
As for traffic control, common ownership is not only a terrible erosion of property rights, it also smells similar to border-security-only "security". "we can implement traffic control with common ownership" sounds similar to "we don't need encryption because we have a firewall". The only thing central ownership gets you is a false sense of security.
It seems sort of unlikely that we will be having fiascos like the jeep hack in 5 years. They mostly mitigated that by correcting the stupid configuration of the connection over the cellular network. It's also easy to imagine that new systems will not trust the messages coming from components that have an outside network connection.
The security thing - that can be fixed. Can't be that hard to isolate the brakes etc from the LTE modem in a sensible manner.
And I also think there will be tangible benefits. e.g. if you think about live traffic routing on Waze etc. That is still pretty damn primitive in terms of interactions that are possible yet already fairly useful.
No more so than securing any other piece of tech connected to the internet & the same principles apply. Split it into zones like you would a corporate network & stick firewalls in between to separate the fluffy online stuff from the core systems. That is ultimately where this is going - a mini network of devices on wheels.
And if something is really too hot to connect at all then you can always isolate that system completely.
Obviously its not going to be 100% fool proof but then again nothing is.
However some car manufacturers don't have separated networks (due to cost reasons).
If every new car is connected, I'm not buying a new car. Simple as that.
How I see this breaking down is that once the automation and connectivity tech has had a chance to be proven statistically safer, the insurance companies will move in.
First, discounts for driving a fully automatic car. Smaller, then larger.
Then, normal cars' insurance gets more expensive as the automatics proliferate. Eventually prohibitively so.
Finally a break point is reached where the legislators see that allowing comparatively death-trap like normal cars on public roads is irresponsible, and they become banned.
Car makers react by eventually ceasing production of manual drive cars by all but special request and for special uses.
Although, a self-driving car would be convenient for overnight driving....
You could be the most attentive, conscientious driver in the entire world - you will eventually make a mistake, possibly a fatal one.
We haven't successfully created bug-free software yet. How exactly is this mythical self-driving car going to pull that off? You did say 'eventually' though, and strictly speaking, that does not put a deadline on these achievements...
I am for self driving cars, as a choice, but those don't have to phone home any data at all to just work fine (in theory).
How all smart cars should work: eg current gen Ford and Lexus cars can use your phone over Bluetooth (if you agreed). And if you habe an accident, the inbuilt 100% offline text-to-speech (Nuance based) solution calls 911 or 112 or <emergency number> and says with a voice similar to Siri exactly where the accident happened (GPS coordinates).
How telco lobby groups think cars should work: cars sends GPS coordinates every minute incl all car telemetry to a central provider using a telco cell data connection - just in case you may have an accident. Well, you can see how it works out if you don't care about your customer and build something no one wants - people bought smaller cheaper cars from Japan and Asia in general. Now Detroit car industry has a big problem.
As a USian who has driven inexpensive cars, something like 2/3 of my car insurance payments have been for medical coverage.
If there is a substantial difference in the medical costs associated with auto drive and manual drive, I'm sure that single payer systems would also account for it somehow.
The point I'm reaching for is that the system where the government pays for medical care will likely respond to a substantial difference in safety, at least in the case it would save the system a substantial amount of money.
In that case, cars with various types of safety systems, including autopilots may have reduced rates. But I'm not sure what would make market insurance rates for manually driven vehicles higher than they are today just because cars with better safety systems are on the road (for which they'll presumably get insurance discounts).
Here's why that won't happen: Motorcycles.
If motorcycles came into existence in, say, the 1970s, they would've never been made road legal. However, there's too many of them out there nowadays that barring their use through legislation would cause a massive backlash. Now apply that to cars, of which there are far more, and which they cost people significantly more money to obtain. Banning them would be political suicide.
Non-automated cars will simply be treated like motorcycles. If you're willing to accept the risks and costs, it's your choice. It just eventually (i.e. 2040 or 2050) won't make sense for the vast majority of people.
So are used cars. A smart person doesn't go out and buy a $20,000 economy car. They go buy a $5,000 used car which in fact will be very reliable when maintained. A lot of people instead spend <= $3K.
People don't buy cheap motorcycles in the US for transportation. I've met plenty of motorcycle owners, of all walks of life, and only have met one or two who bought it as their only and primary transportation.
>then getting regulated out of existence
Good luck getting campaign support from Porsche, Ferrari, Lamborghini and other exotic and sports car (ex. there are more Corvette owners than Porsche sports car owners) owners on that one.
There's no backlash against banning Toyota Corollas. However the die-hard car enthusiasts will fight tooth and nail to ensure their expensive sports cars aren't useless, and there's enough of them (with enough money) to ensure any such plan to keep them off the road will fail (and to ensure you never get elected to office again).
The "problem" will mostly solve itself, as eventually (decades), TCO alone will result in most people not owning and operating their own vehicle. Most of what will be left are simply enthusiasts, of which there are a significant enough number (same with motorcycles). They'll pay what it costs to enjoy their toy, while the remaining 99%+ of the population won't care for such things and rely on autonomous cars. Moreover, autonomous cars are going to have decades of development having learned to deal with non-autonomous cars.