Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

The ethics and business implication of an 8k $ Autopilot

This site may earn commission on affiliate links.
1. Tesla is using Autopilot as a option aimed at increasing margins and by grossly overpricing it, Tesla severely limits access to it.
Elon Musk , Google and pretty much anyone with a clue are clear, the robot increases safety. For level 1 and 2 one can argue that it is a convenience feature but level 3 to 5 are absolutely a safety feature. A prohibitive price for a safety feature is clearly unethical, especially when they could do better with little effort - even if it would have been difficult, Tesla isn't afraid of difficult tasks.That'st sets them apart from everybody else.
From an ethics perspective Tesla should include the feature by default and make every effort they can to encourage customers to use it. It's safer for the driver and every other participant in the traffic.
So you are claiming that Tesla has an ethical obligation to include full autonomous driving capability as a standard feature at little or no extra cost. You apparently have no idea of the development and implementation cost of that feature, which does not yet exist in a Tesla car or in any other car for sale in the world today. It doesn't yet exist because it is an extremely hard problem to solve and even Elon admits it is years away (in numerous interviews over the past year), not months away. You say Tesla "could do better with little effort". You have no idea what you are talking about, and You are holding Tesla to an absurdly high standard that is divorced from reality.
 
Had to make an account as i really don't like the ethics and financials of charging 8k$ for Autopilot.

Tesla and Elon Musk must ask a few questions.
1. Is it ethical to severely restrict access to a safety feature?
1. Can we make it available to everybody without losing revenue and margins?
3. Are there any additional upsides or downsides when including the feature in the base model.

1. Tesla is using Autopilot as a option aimed at increasing margins and by grossly overpricing it, Tesla severely limits access to it.
Elon Musk , Google and pretty much anyone with a clue are clear, the robot increases safety. For level 1 and 2 one can argue that it is a convenience feature but level 3 to 5 are absolutely a safety feature. A prohibitive price for a safety feature is clearly unethical, especially when they could do better with little effort - even if it would have been difficult, Tesla isn't afraid of difficult tasks.That's what sets them apart from everybody else.
From an ethics perspective Tesla should include the feature by default and make every effort they can to encourage customers to use it. It's safer for the driver and every other participant in the traffic.

2. Model 3 is a very costly car, even in the US it is a stretch to call it affordable.Elsewhere 35k$ means even more. Sure, folks will take advantage of gov grants, maybe do the math for gas savings and ,hopefully, lower maintenance and repair costs and , best case scenario, conclude that Model 3 is getting closer to a 20k$ ICE car. When you buy a 20k$ car, adding 8k$ is more than too much.So it would be reasonable to conclude that only some 10-20$ of the global Model 3 orders will include Autopilot. Limiting access like this when everybody has the hardware only increases the pricing for the few that can afford to buy it.
If only 10-20% of buyers can afford the feature at 8k$, it would mean that to include the feature in the base model, thus allowing access to every buyer, would lead to a price increase of 800 to 1600$ to be able to maintain the same overall revenue. Fine tuning overall costs and adjusting pricing for other options could allow for the feature to be included in the base model without a price increase while keeping Tesla's revenues at the same levels.- do remember that almost any feature in the base model of any car is a CHOICE made by the car maker. the car maker decides what is important enough to make it in the base model and , in this case, despite claiming that it is a crucial safety feature, Tesla has decided not to include it in the base model.
It's like charging 2k$ to enable the airbags and , as i have explained, it's not something they need to do, it's something they chose to do.


3.
Upsides:
- including it in the base price boosts sales and is great marketing
- keeping the drivers alive helps a bit as dead people don't spend money
- provides more differentiation and increases customer retention
- for car as a service to ramp fast, people need to trust the robot. Allowing more consumers to experience autonomous features ahead of such a service, speeds up the adoption for car as a service and that's worth tens of billions, maybe hundreds over the next 10-15 years
- if in the early stages of car as a service owners also send their cars to earn money, fewer cars with the feature leads to fewer cars available for the service and the entire point of this, from Tesla's perspective, is to enable the growth of the service much quicker and without the CAPEX required for their own fleet.
- it's easier to argue that it's a crucial safety feature (and accuse the press of killing people :p) if you don't charge for Autopilot more than what 80% of the people on this planet earn in a year
- autopilot would also lead to lower under warranty costs for Tesla as the robot is a smoother operator (lol)
- customers are more likely to enable data sharing

Downsides:
With more autonomous cars ,negative publicity and legal complications would increase if the system is less than stellar.

For X and S the rate of adoption will be higher but it's also possible that the price of Autopilot will be lower for Model 3 and if that's true, the math would become even easier. It's also possible that including the Autopilot feature in the base price might require a slight increase in price but adding 500$ to the base price this late in the game when the design is finalized and it's harder to cut costs elsewhere but would be something that the vast majority of customers would welcome and the ethics would be on Tesla's side- as long as it is a minimal increase.

It's easy to do things like everybody else without giving it any thought at all but that doesn't mean that it is good business or .. decent.
However, Tesla has the opportunity to do better here. Offer a better car, a safer car to every customer while making more money and paving the way for a fast adoption for a car as a service offering.
Tesla is all about doing it better than others and finding ways to do so.It's easy to do much better here and not doing better kills people.


PS: I understand that most people here will feel the need to defend Tesla's current position at any cost instead of trying to understand my point of view and maybe accepting that Tesla could do better, much better. Defending the status quo without seriously considering a better way to do something is almost always a bad idea. Tesla exists only because a few people decided to build a different car , a better car and rejected the status quo. That was much harder than this, this would be easy.
Is it ethical to charge for selling cars. Or how about charging for food. If you cannot pay for food you will die
 
Since all cars will have the hardware, I wonder if Tesla will allow cars that haven't purchased self driving to participate in the Tesla ride sharing system. You could then "earn" your self driving feature for personal use. Tesla could even set up a system where you don't get the cash profits, instead they're applied to new features.
 
  • Like
Reactions: brianman
I bet my life that a driver with enough sleep, experience behind the wheel, full attention to driving, well maintained vehicle with a defensive driving mindset is safer than AP2.0 or even AP 5.0 whenever that decides to come out. I've driven a car without autonomy for 20 years so in my mindset, even AP 2.0 is a toy that needs to be approached very cautiously.

Go cross post on gm-volt and complain about the LT Trim Volt not having access to any of the Driver Confidence packages.
 
  • Disagree
Reactions: Yuri_G
I'm not sure a US$43k base price Model 3 with full AP standard would have quite the appeal of a US$35k base price Model 3 electric car. But with AP optional you have the choice to include it if you think it is critically important. It does look like an AP car and a non-AP car will have the same safety when driven manually. That should be fine for now.
 
I think you're underestimating the difficulty of the problem. Building an electric car was trivial by comparison. It didn't require any new inventions, just really good execution of existing technology. By contrast, full autonomy is an unsolved problem. There are many prototypes that can handle simple tasks, but think about the most complex intersection in the middle of Boston, or a ladder that falls off a pick up truck in the middle of the freeway. That all has to work to claim full autonomy.

And that's not even getting into the ethical questions, like if the car is in self driving mode, and it sees two little girls run into the street, and the only way to save their lives is to veer into a telephone pole, costing your own life, which is the correct choice?
"... two little girls run into the street ..." obvioulsy a city or suburban street, so speed is probably 25 to 35 MPH which implies you have very good odds to survive crashing into a tree, IF you are in a Tesla.
Remember, these (S/X) may be the safest cars ever built.

Safety (from Tesla website)
Model S is designed from the ground up to be the safest car on the road. Much of its safety is owed to the unique electric drivetrain that sits beneath the car's aluminum occupant cell in its own subframe. This unique positioning lowers the car's center of gravity, which improves handling and minimizes rollover risk, and replaces the heavy engine block with impact absorbing boron steel rails.

Side impacts are met by aluminum pillars reinforced with steel rails to reduce intrusion, protecting occupants and the battery pack while improving roof stiffness. In the event of an accident, eight airbags protect front and rear occupants, and the high voltage power source is automatically disconnected.
  • Eight airbags for driver and passenger heads, knees, and pelvis plus two side curtain airbags
  • Electronic stability and traction control
  • Four wheel antilock disc brakes
 
And that's not even getting into the ethical questions, like if the car is in self driving mode, and it sees two little girls run into the street, and the only way to save their lives is to veer into a telephone pole, costing your own life, which is the correct choice?

Speaking as someone with a degree in philosophy - nobody cares about that ethics 101 thought experiment (there's a trolley comin' down the tracks - who do we kill? what if there's more on that side, or this side? What if it's your Great Aunt Hilda? blah blah blah).

There is too much money to be made - the insurance industry and auto industry will work out the liability details and the lap dog regulators will get it done. The liability risk will be amortized over millions of cars. You're making a mountain out of a molehill - it's a nice problem for bloggers and us hobbyists to wring our hands about but it's tiny potatoes in the big picture.

Peeps get killed - sh*t happens. Payouts happen. Billions are made. The beat goes on baby. . .
 
Had to make an account as i really don't like the ethics and financials of charging 8k$ for Autopilot.

Tesla and Elon Musk must ask a few questions.
1. Is it ethical to severely restrict access to a safety feature?
1. Can we make it available to everybody without losing revenue and margins?
3. Are there any additional upsides or downsides when including the feature in the base model.

1. Tesla is using Autopilot as a option aimed at increasing margins and by grossly overpricing it, Tesla severely limits access to it.
Elon Musk , Google and pretty much anyone with a clue are clear, the robot increases safety. For level 1 and 2 one can argue that it is a convenience feature but level 3 to 5 are absolutely a safety feature. A prohibitive price for a safety feature is clearly unethical, especially when they could do better with little effort - even if it would have been difficult, Tesla isn't afraid of difficult tasks.That's what sets them apart from everybody else.
From an ethics perspective Tesla should include the feature by default and make every effort they can to encourage customers to use it. It's safer for the driver and every other participant in the traffic.

2. Model 3 is a very costly car, even in the US it is a stretch to call it affordable.Elsewhere 35k$ means even more. Sure, folks will take advantage of gov grants, maybe do the math for gas savings and ,hopefully, lower maintenance and repair costs and , best case scenario, conclude that Model 3 is getting closer to a 20k$ ICE car. When you buy a 20k$ car, adding 8k$ is more than too much.So it would be reasonable to conclude that only some 10-20$ of the global Model 3 orders will include Autopilot. Limiting access like this when everybody has the hardware only increases the pricing for the few that can afford to buy it.
If only 10-20% of buyers can afford the feature at 8k$, it would mean that to include the feature in the base model, thus allowing access to every buyer, would lead to a price increase of 800 to 1600$ to be able to maintain the same overall revenue. Fine tuning overall costs and adjusting pricing for other options could allow for the feature to be included in the base model without a price increase while keeping Tesla's revenues at the same levels.- do remember that almost any feature in the base model of any car is a CHOICE made by the car maker. the car maker decides what is important enough to make it in the base model and , in this case, despite claiming that it is a crucial safety feature, Tesla has decided not to include it in the base model.
It's like charging 2k$ to enable the airbags and , as i have explained, it's not something they need to do, it's something they chose to do.


3.
Upsides:
- including it in the base price boosts sales and is great marketing
- keeping the drivers alive helps a bit as dead people don't spend money
- provides more differentiation and increases customer retention
- for car as a service to ramp fast, people need to trust the robot. Allowing more consumers to experience autonomous features ahead of such a service, speeds up the adoption for car as a service and that's worth tens of billions, maybe hundreds over the next 10-15 years
- if in the early stages of car as a service owners also send their cars to earn money, fewer cars with the feature leads to fewer cars available for the service and the entire point of this, from Tesla's perspective, is to enable the growth of the service much quicker and without the CAPEX required for their own fleet.
- it's easier to argue that it's a crucial safety feature (and accuse the press of killing people :p) if you don't charge for Autopilot more than what 80% of the people on this planet earn in a year
- autopilot would also lead to lower under warranty costs for Tesla as the robot is a smoother operator (lol)
- customers are more likely to enable data sharing

Downsides:
With more autonomous cars ,negative publicity and legal complications would increase if the system is less than stellar.

For X and S the rate of adoption will be higher but it's also possible that the price of Autopilot will be lower for Model 3 and if that's true, the math would become even easier. It's also possible that including the Autopilot feature in the base price might require a slight increase in price but adding 500$ to the base price this late in the game when the design is finalized and it's harder to cut costs elsewhere but would be something that the vast majority of customers would welcome and the ethics would be on Tesla's side- as long as it is a minimal increase.

It's easy to do things like everybody else without giving it any thought at all but that doesn't mean that it is good business or .. decent.
However, Tesla has the opportunity to do better here. Offer a better car, a safer car to every customer while making more money and paving the way for a fast adoption for a car as a service offering.
Tesla is all about doing it better than others and finding ways to do so.It's easy to do much better here and not doing better kills people.


PS: I understand that most people here will feel the need to defend Tesla's current position at any cost instead of trying to understand my point of view and maybe accepting that Tesla could do better, much better. Defending the status quo without seriously considering a better way to do something is almost always a bad idea. Tesla exists only because a few people decided to build a different car , a better car and rejected the status quo. That was much harder than this, this would be easy.
PS- "... Autopilot more than what 80% of the people on this planet earn in a year ..."
Yes, and 80% of the planet probably shouldn't even bother buying cars. Walking, bikes, public transport and you'll save $5-10,000 in auto costs (see AAA for cost numbers).

We may have reached peak car Peak car - Wikipedia and I have seen car mania all over the world. For example on Westmen (Vestmannaeyjar) Islands, Iceland. Less than 12 sq Km and it took about 20 min. to walk across the the entire town (~5,000 population) a fishing village. They had an airport, and a harbor. Fisherman in 1972 making good money and everyone on the island had cars. What else to spend money on? In spite of a 100% import duty on cars/parts/tires. (SUV/trucks no import duty, can you guess why? trying to limit demands for paved roads). I asked people what did they use their cars for (no commutes, you could literally walk everywhere and cars had to stay on roads).
Answer: We drive down to the harbor (10 min. walk at most) and listen to the car radio. Seriously.
 
  • Informative
Reactions: gabeincal
People respond to incentives.

If you incentivize them to invent... they will invent.
If you incentivize them to be lazy panhandlers... they will be lazy pan handlers.

Feel free to tell the intelligent, driven, creative people that they should consider giving away their life's work, and ... they will be incentivized to stop creating all the wonderful things that socialists love to claim as their own.

Cheers!
 
Tesla has lots I disagree with, but only part I find truly indefensible is the unpredictable deprecation. They knew ab new AP when I bought my car 2 weeks ago and asked them repeatly about it. They knew telling me the truth would cost them a Q3 sale. I'm ok with it. I didn't buy a $100,000 car because it would go up in price, but what you are saying makes no sense as is not American/capitalist.
If it wasn't what you wanted, why did you buy it? You asked story employees and I'm sure they didn't know. It will take time for this new hardware to get the new software before the promises of autonomous driving is realized. You just saved yourself from being a beta tester. Get a new Tesla when all the features are flushed out. And you'll have more choices of vehicles to choose from.

Enjoy your new ride.
 
Your post is based on two critical assumptions: you believe this self-driving problem is easy to solve and that the overall cost to solve it (both the research/development and the actual hardware/software for each car) is low, such that Tesla could have chosen to include this technology into all cars with little or no additional cost to the buyer. From these assumptions, about which you have no factual information at all, you deduce that Tesla is "grossly overpricing" this technology, and that the $8,000 cost is "not something they need to do, it's something they chose to do."

Your logic is reasonable, but your logic rests blindly on the aforementioned two assumptions, and those are not reasonable. If your assumptions are wrong, and I believe they are, then the most perfect logic will lead you directly to a wrong conclusion. Show me more information (as in real, verifiable data) about why you believe those two assumptions are true, and then we'll talk about the ethics of the choices Tesla has made.
 
Had to make an account as i really don't like the ethics and financials of charging 8k$ for Autopilot.

Tesla and Elon Musk must ask a few questions.
1. Is it ethical to severely restrict access to a safety feature?
1. Can we make it available to everybody without losing revenue and margins?
...A prohibitive price for a safety feature is clearly unethical, especially when they could do better with little effort - even if it would have been difficult, Tesla isn't afraid of difficult tasks.That's what sets them apart from everybody else.
From an ethics perspective Tesla should include the feature by default and make every effort they can to encourage customers to use it. It's safer for the driver and every other participant in the traffic.

I 100% agree! No car in the world should be sold until they have these features. Since no company has been able to create this technology they must all license it from Tesla.

Would you rather Tesla increase the price of each vehicle $8,000 and make this standard and sell less vehicles?
 
Speaking as someone with a degree in philosophy - nobody cares about that ethics 101 thought experiment

Speaking as an engineer who designs complex systems -- we don't have the luxury of not caring. Software design consists of two main activities: architecture (the fun part), and an endless series of what-ifs called implementation. All of them have to be addressed, or the software has undefined behavior, which is unacceptable in a highly regulated, life-critical system.

If you don't like the philosophical angle, just consider the practical issue that driving legally and driving safely are not always compatible. If there's an enormous pothole in the road, and the car can safely avoid it by briefly veering over a double yellow (no oncoming traffic at the time), it should do that. It's easy to list hundreds of examples of like this. And of course it's possible to solve them all -- it's just an enormous amount of hard, tedious, time-consuming work.
 
  • Informative
Reactions: number12
... think about the most complex intersection in the middle of Boston, or a ladder that falls off a pick up truck in the middle of the freeway. That all has to work to claim full autonomy.

What's your definition of "work?" My definition is that the computer can do the job with sufficient reliability to be acceptable to society. What level is acceptable? Now you're getting into the realm of arbitrary value judgments. To you, maybe you need a system at least as reliable as a human in all conceivable situations. I am not that picky myself. If you give me overall system performance almost as good as a human I will be happy because I am gaining one helluva lot of utility (my time) in return for some amount of increased risk. Everything is a tradeoff.

And fine - since you brought them up - let's look at the most complex scenarios. I have two things to say:

1 - Humans are not perfect in the most complex scenarios so why expect computers to be (especially in the early days)? A ladder falls of a truck in front of you, ok. Many human drivers would hit the ladder. Sh*t happens. We are not perfect. Sometimes there is no acceptable solution but to run over the ladder, or physics won't let you stop in time. Etc. So - what's the conclusion here?

2 - Complex Boston intersections. Just because we cannot yet imagine the solution to a computer interacting with a complex system does not mean it does not exist. There is a video from the last six months of Amnon Shashua showing Mobileye's neural networks learning to navigate through roundabouts with unpredictable human drivers. The networks learned, dude. It was unnerving watching that video.

Speaking as an engineer who designs complex systems -- we don't have the luxury of not caring. Software design consists of two main activities: architecture (the fun part), and an endless series of what-ifs called implementation. All of them have to be addressed, or the software has undefined behavior, which is unacceptable in a highly regulated, life-critical system.

If you don't like the philosophical angle, just consider the practical issue that driving legally and driving safely are not always compatible. If there's an enormous pothole in the road, and the car can safely avoid it by briefly veering over a double yellow (no oncoming traffic at the time), it should do that. It's easy to list hundreds of examples of like this. And of course it's possible to solve them all -- it's just an enormous amount of hard, tedious, time-consuming work.

You make good points and "who cares" was flippant language on my part. I take it back - let me rephrase more carefully my point. I am not saying engineers should not work on the problem of saving humans. I am saying that no reasonable person expects you engineers to pull it off with 100% perfection. OF COURSE the software will have undefined behavior - and of course it is acceptable as well! The universe has much behavior we cannot define. You engineers cannot model every scenario - especially with a software based neural network which is not hand coded. Even the computer scientists do not know the specifics of why some AI learning algorithms work better than others. You can never predict with 100% accuracy what output a neural network will give you for a given input. It isn't a hand coded logic system.

But so what? They learn, things get better over time, everything will work out in the end. Do you *really* think there are not detailed planning sessions going on behind closed doors right now between high level insurance execs, auto execs, industry regulators and legislators? Of course there are.

Then the marching orders are given to the engineers, and unpredictable behavior is insured. Everything will be fine.
 
  • Like
Reactions: flashflood