Everybody hates tesla, Part 2


The following is a transcript of the audio available via the player above.  The audio file is the definitive source.

Welcome, welcome, welcome everyone to this episode of Tech. Cars. Machines. My name is Ali Tabibian. You have everything you need in the episode notes, so let's get moving.

This is part two of Everybody Hates Tesla. In part one we talked about how electrification of vehicles, the EV, the electrical vehicle is going to change things for the original equipment manufacturers, the Toyotas and the General Motors of the world.

We talked a little bit about why we think electrification will change the industry's profit structure well ahead of its unit volume build-up. Take a listen, it's pretty interesting.

Because we looked at the world of electrification through a Tesla lens there are a few other changes that Tesla is bringing about to the automotive world which we're wrapping up under the same heading. In this part two we're going to discuss three things.

One is, what does electrification do to the automotive suppliers, the people who feed into the General Motors and the Toyotas of the world? We're going to talk about dealerships and the dealership structure in the United States and how Tesla is disrupting that traditional way of selling cars in this country, and of course we'll talk a little bit about self-driving and how Tesla has probably unnecessarily tried to make both claims in that environment.

Voiceover:                   01:16               Tech. Cars. Machines. Subscribe here or at gtkpartners.com.

Ali Tabibian:                 01:23               First, let's talk about the supply chain. Electric vehicles are going to be a challenge for the automotive supply chain as that chain is currently structured. Let's dig into what happens inside a typical combustion engine vehicle to find out why.

When you really think about what's happening in a combustion engine it's really incredible what a fiery, loud and ferocious set of events are involved and what a miracle it really all is. 

A gas engine which is formerly called an internal combustion engine or ICE for short, when you think about it, really what a miracle. I mean Rome and Paris, two cities that have been around oh what? 2,000, 3,000 years, for almost all of their history a trip from one to the other was probably a once in a lifetime event for almost everyone.

Today with 30 gallons, about 120 liters of a thin liquid called gasoline or petrol, you can take a trip between one and the other ... you can take a trip between them, between those two cities of Rome and Paris in about 14 hours. 

Think about that, 30 gallons is less than what two Americans use for showers in the morning. For that amount of liquid you can transport thousands of pounds of a vehicle at high speeds over 1,000 miles.

This elegance hides what a large number of mainly mechanical systems are required to make this all happen. As we take you through what's happening under the hood I'd like you to stay focused on and think about how much of what we're describing is or is not necessary in an electric vehicle?

Let's start with the fuel. Fuel systems take the gas from a tank in the back of the car to the front where the engine is. Some systems aerosol the gasoline and mix it with air and this vapor is then injected into a metallic chamber, it's called a cylinder, where a spark plug basically causes it to explode, and I do mean explode.

I mean the temperature and the pressures inside a cylinder are around 600 degrees Celsius, which is around 1,100 degree Fahrenheit, with 3,250 times the atmospheric pressure. This pressure is what forces down a metallic block called the piston that then slides down inside the cylinder. This happens 30 to 60 times a second per cylinder. Think about that, per second, double that rate if you're really gunning the engine. 

Just imagine how precise the manufacturing has to be, how special the lubricants have to be, how good the heat evacuation systems have to be, how good the noise and the vibration isolation has to be, just to keep this bit of the system working on a sustainable basis for hours on end. It's amazing, really amazing. After all this you have a piston that's basically moving up and down which is not that useful, so you need a series of mechanical systems that take this up and down movement, convert it to rotational motion, and deliver it to the car's wheels.

These are called the driveline systems. This process starts with a piece of metal, it's called a camshaft, that both combines the motions of all the cylinders and turns it into a rotational motion, it takes that up and down and turns it into rotational motion, and it typically does this underneath the engine block.

From there other metal pieces basically called the powertrain system transfer this movement, changes its access to line up with the center of the relevant wheels and deliver the power there.

Sometimes these mechanisms have to also reset the compromise between power and speed that the engine is delivering, and that's why you need a transmission system. That's what the transmission is for. Sometimes the relevant wheels are in the front, a front-wheel drive system, sometimes you need to get the power to the back, which is a rear-wheel drive system. All of this requires big, heavy mechanical, generally metallic components.

All of these systems by the way have to last under tremendous punishment. For example, the camshaft I mentioned probably makes about 500 million rotations over a 10 year period of driving. We also leave our cars outside under the sun, sleet, rain, drive them over punishing potholes, how long would you last if you were treated this way?

After all this an internal combustion engine vehicle also has to worry about the exhaust from that explosion. You have to take that exhaust from the front of the car where the engine is, to the back where the exhaust pipes are and in the process you have to scrub it in what's called a catalytic converter, and then increasingly with frightening consequences blow it out into the atmosphere, the same atmosphere we use to breathe. So the whole cycle converts a liquid gasoline to some motion, a lot of waste heat, and some waste gas coming out of the exhaust pipe.

Now let's compare this to what happens in an electric vehicle. In the EV case instead of a gas tank your energy is stored in a battery pack, in a chemistry that converts it to a usable form, mainly electricity, without all the commotion of a combustion engine.

The battery does need a cooling system not any more complicated than what you'd need for a combustion engine's heat, and the battery also needs very sophisticated management of your software, and it can be very heavy. Tesla's Model S battery weighs 1,000 pounds. 

The energy from this battery using some normal looking wires is very simply delivered to wherever you want it to be, front, back, all four wheels, wherever you want. The engines for an electric vehicle are called motors, electric motors.

They're pretty small, about the size of a basketball in the case again of a Model S, and therefore they can be squeezed in fairly easily wherever you need them to be, say right next to the rear wheels in a standard Tesla or next to the rear end of front wheels in the case of a four-wheel drive version. Of course there's no exhaust worry about, so you can throw away or forget about the exhaust systems and the catalytic converter that we talked about earlier.

What's also cute about these motors is that they are two-way or symmetric, meaning that in the same way that can use electricity to turn, they can take a turning motion and produce electricity. For example, when you're traveling in an electric vehicle and you take your foot off the accelerator you cut off electricity to the motor, but the car obviously has some momentum because it's moving, so the wheels are turning, which in turn forces the electric motor that are connected to them to keep turning even though they are technically, "Off."

This forced motion produces electricity and recharges the battery. This is kind of cute and this contributes to why electric vehicles are way more efficient than ICE engines or internal combustion engines in highway stop and go traffic where you're going from 50 to 10, back to 50 all the time.

So far what we've described of course is pretty dramatic, it's a pretty dramatic set of the differences between ICE and EV, but a lot of it is probably been somewhat intuitive to you. Electrification tends to result in huge declines in mechanical complexity, and you have plenty of examples around you in the world.

Just think about going from hard disk drives to solid state drives, and what's that done for your iPod versus your iPhone, as well as your laptops? But one EV feature that really is important economically to electric vehicles, but is less intuitive is the following, EVs have inherently good acceleration from a standing start, something consumers are willing to pay for and at least in this country, in the United States, really appreciate.

Colloquially this is the 0-60 acceleration of a vehicle, the reason it's listed every time you look at a vehicle specification is because the more you have, the more people will pay for that vehicle.

I really want you to understand why this is from a fundamental point of view, so let's geek out for a second. Why do electric motors give you more immediate acceleration than ICE?

For an internal combustion engine to generate power it needs its pump to suck in that air and fuel vapor that we talked about, but that pump runs on the engine's power itself, so there's a chicken and egg problem, you can get in the air and fuel into the engine until the engine is turning at a good clip, and at idle or at a standing stop it's not turning very fast.

So you need to sort of go through this chicken and egg back and forth for a little while until you build up power and then you get that surge of acceleration, the engine opens up as gearheads like to say and you get that opening up feeling when the car really starts to get that surge of acceleration.

Also remember the powertrain or the drivetrain that we talked about a couple minutes ago in a combustion engine, well that's basically a lot of weight that the engine has to get moving to get its power from the engine to the wheels. There's an inertia that's slowing down the build-up of rotations per minute of the engine and therefore of the engine's ability to produce power, but once you get everything turning then the ICE really starts to sing and you feel that surge of power delivery and you get that surge of acceleration that pushes you back into your seat.

Electric motors do in fact have a type of inertia but it shows up late in the acceleration cycle rather than early. ICE, when you accelerate you're fighting against that inertia right away, in electric motors actually this inertia shows up when you've increased your rotations per minute.

Here's why. Remember we said an electric motor where you mechanically force it to rotate can generate electricity. It turns out when you feed the motor electricity to get it rotating that motion that you're creating with electricity eventually generates a bit of negative electricity called the back electromotive force or back EMF. 

                                                            So you start the motor turning with electricity, it quickly spins up, it generates back EMF, which makes the motor less efficient. This is better behavior than an ICE engine because the consumer feels or the driver feels a quick acceleration when they press the pedal. You also don't need a transmission to help you generate more torque when you're starting the engine off. 

                                                            By the way in electric vehicle you don't even need a transition to run the car in reverse, you don't need to change gears, all you need to do is switch the negative and positive polarity of the battery connection, reverse its polarity as it's called, and the engine will run in reverse and so will the car. Theoretically an electric vehicle is just as fast in reverse as it is going forward. 

                                                            Here's the message again. There are a lot fewer parts in an electric vehicle, no fuel systems powertrains, drivetrains, exhaust systems. Each of these names, each of these phrases refers to a major business unit at each of the many auto suppliers out there which are probably employing around three million people around the world.

                                                            Now there's plenty of room for growth for suppliers to participate in electric vehicle components, the battery management systems and so on, but at this point electric vehicle share growth is probably a net loss to and if not that clearly a strong challenge and substantially discontinuous with the skills of the massive auto supply industry as is currently arranged.

                                                            Oddly enough, it's the suppliers of the simple stuff, the heating and the venting, and the air conditioning, the infotainment, the wiper washers, et cetera, it's those suppliers whose work really doesn't change that much in the realm of electric vehicles.

                                                            Here's a cute anecdote in the case of self- driving vehicles with a lot of cameras and a lot of lidars, which need to shoot out a beam of light and receive it back, what's actually interesting is the wiper washer people, the people who know how to keep those lenses clean are in fact seeing an increase in demand for their systems.

                                                            These type of issues by the way affect the suppliers disproportionately more than the original equipment manufacturer. So they affect the Delphies of the world which feed into the General Motors and the Toyotas of the world more than GM and Toyota themselves.

                                                            The reason for that is that you can think of the OEM, the GM and Toyota as mainly an assembler of a vehicle. They make some of their parts themselves, they frequently make the engines themselves, but beyond that they stamp the metal and then they get parts from everybody else and plug it into this metal container and put on top of the chassis, which they frequently also design themselves.

                                                            And so, when those internal components change all they do is pass on that change to whoever's supplying them those components, and if those components are no longer needed that's not a big deal for GM and Toyota, but it is a big deal who are supplying it into those vehicles.

                                                            So that's where you'd expect a lot of the change and the disruption in the employment, maybe employment loss, but certainly employment change to take place. By the way, our point isn't to say that the days for suppliers are numbered, Tesla's experience with the unreliability of a lot of its vehicles and the scaling problems it's had with production show you how well the current configuration of the industry actually works, and how good the suppliers are working with the OEMs and delivering the appropriate parts at the appropriate time.

                                                            Think about all the production problems the Model 3, the midsize Tesla has had and everyone including Tesla would have probably been better off if they had just handed over the manufacturing of that vehicle to suppliers like Magna, who also for example manufacture completely certain BMW 5 series variants. There are enormous skills that the suppliers have built up, but just like the profit share distribution argument that we made in part one there's a shakeup in the offing. 

                                                            Tesla is even a more specific threat to the automotive supply chain. A number of us at GTK were treated to a tour at SpaceX, which is Elon Musk's space launch company, as I'm sure many of you know. We were struck by how aggressive the insourcing was at that Elon Musk company. Tesla seems to be on the same path. 

                                                            In addition to building its own batteries Tesla is also departing from automotive norms by building, let's see, seats, robots, cameras and even vision chips. Most auto companies can't imagine engaging in any of these activities. Look at the featureless dashboard of the Model 3, and which are seeing their suppliers losing out.

                                                            The dashboard itself is coming from a supplier but the second tier suppliers which would typically supply knobs and gears, and dials, and whatnot, are basically, have no role. Every feature in that dashboard is embedded in software inside a single screen in the middle of the car.

                                                            Here's another thing to keep in mind. Tesla which has had a history of reliability problems also has a habit of publicly blaming suppliers for airbag, door and manufacturing problems when things go wrong. This is another reason for the supply chain to dislike the company, and this is a problem.

                                                            A substantial amount of innovation in the automotive industry is still coming through the supply chain, and you can imagine with all the factors we've discussed that that supply chain would have a particular incentive for Tesla's competition to win out over that company. This would be a concern for the company's investors if we were them. 

Voiceover:                   16:27               Tech, fancy electric cars. Machines, click subscribe here and visit us at Tech and gtkpartners.com.

Ali Tabibian:                 16:37               Now let's talk about dealerships and software upgrades, or in other words circumventing 1.1 million jobs in the United States. You'll be surprised to know that your legislator demands you buy your car from a dealership independent of whoever made the car.

                                                            It's actually true. All 50 states have some sort of franchise laws as they're called. These laws at a minimum prohibit, direct to consumer sales of new vehicles from a car lot. Occasionally they'll go further by protecting dealers from competition in "Relevant market areas." Sometimes they'll even prevent the auto manufacturer from servicing the vehicle directly. 

                                                            Why is this the case? It comes back to that 1.1 million jobs that we opened this little segment with, the dealership lobby's power comes from two sources. At the statehouse level it's a very large source of sales tax revenue and substantial employee base. Nationally the dealerships help the auto manufacturers manage revenue and inventory.

                                                            For example, an automaker considers a vehicle sold when it ships it and it arrives at the dealer lot, not when the dealer sells it. The dealership lobby justifies its privileged status with the uniquely virtuous local citizen strategy. This is actually a term of art, if you will uniquely virtuous local citizen, which applies to a lot of mercantilist approaches that certain industries take, they're trying to have essentially legislative protection for their way of doing business.

                                                            What they do is highlight consumer benefits in terms of service, employment, charitable contributions, you know, the proverbial Little League baseball team that these auto dealerships sponsor, these are all arguments for why they should be treated differently under the law.

                                                            Now of course, this is nonsense. The Apple stores direct to consumer model, as they are, deliver sales tax, employment and other local benefits as well without needing any regulatory favoritism. 

                                                            Like successful iconoclasts before it, PayPal Uber, Airbnb come to mind, Tesla is challenging these legacy protections not by flaunting the law directly but by operating in a legal gray area. It has, "Showrooms," not, "Dealerships," which demonstrate models but relegate the actual purchase let's say, to an in-store computer. These, "Showrooms," carry no inventory and therefore try to get around the dealership rules. 

                                                            In parallel now the company is fighting a legitimate state by state legal battle with mixed results really. In California Tesla of course has free rein. In New Jersey it's allowed four dealerships. In Michigan not even a service center. Take a guess as to why Michigan has been one of the harder state nuts to crack?

                                                            Tesla's managed to achieve reasonable scale with this approach, 100,000 vehicle sales a year. Basically what this allows the company to do is to threaten the decades old channel structure of the new vehicle industry, that remember employs a little over a million people.

                                                            Now again we're not saying independent dealerships are bad or necessarily even the wrong model, what's actually quite notable is that like many locally owned businesses these dealerships do have superior customer service.

                                                            For example, our own anecdotal researched, but still anecdotal experience is that you are much more likely to find uninformed and disinterested staff at a Tesla showroom than the comparable luxury dealerships in the same area. This goes for our samples in Northern and Southern California.

                                                            What we are saying though is that Tesla is showing its direct to customer model to be effective and something that has not been experienced in this country for about 80 years.

                                                            This demarcation between the auto manufacturer and consumers has led to a number of interesting side effects which once again Tesla is serving as an iconoclast toward. The auto manufacturers were never in the mindset of directly accessing the vehicle because for the most part they never really directly service it.

                                                            What this means is that before Tesla no one upgraded vehicle software over-the-air, a commonplace occurrence by the way in all manner of consumer devices. We've all been delighted and occasionally frustrated by the typical over-the-air updates that come let's say for your iPhone. This doesn't happen in the automobile industry, your fancy car will have the same annoying, quirky software behavior year after year, after year, and for the most part will have the same out of date maps for its navigation system year after year, after year, unless you go out and buy a CD or go to the dealership to get to a new map burned in.

                                                            Keep in mind when a software update becomes necessary or available the OEM pays the dealer to take in the vehicle and manually update the software. In fact there are agreements between these OEMs and the dealer network that frequently prohibit and run around the dealership of which an over-the-air software update can be considered.

                                                            As you can imagine this imposes a lot of costs and it's a major cost disadvantage for the typical OEM versus Tesla. The dealerships of course have an incentive to continue preventing these over-the-air updates because of course it cuts them out of the process and when you go to the dealership you might buy something else or even get have them put you in a new car.

                                                            Here's a note from Tier 1 supplier Delphi, "Right now the industry has a very antiquated way of dealing with software fixes. It goes to a dealer, it gets re-flashed and the dealer gets money." That's Delphi Chief Financial Officer Joe Massaro during an investor day last September.

                                                            In the future when carmakers want to push through a software fix, "We'll turn it on and fix it over-the-air, and if they don't want us to do that for whatever reason then we're not paying for it."

                                                            Tessa's ability to over-the-air almost every vehicle function is years ahead of others, and potentially it makes it the vehicle of choice for delivering alternative transportation models because those models depend on specific measurements that the car makes and delivers to some third party.

                                                            Tesla essentially invented this over-the-air capability beginning with the Model S. Compare this with General Motors who says maybe by 2020 they'll have the ability to regularly do over-the-air updates just for the infotainment system. Tesla meanwhile can send an over-the-air update, change your braking distance and raise or lower the chassis of the vehicle.

Voiceover:                   23:06               Haven't we made you feel like really smart? Let us elevate you to very stable genius. Click subscribe. Subscribe with the little button in your podcast app or click the three dots in the little circle or visit us at gtkpartners.com.

Ali Tabibian:                 23:21               Last and really so far least from our perspective let's talk about self-driving a little bit. You've noticed we've generally been arguing that Tesla's influence on the world of automobiles is underappreciated as far as the fundamentals go.

                                                            We tend to have the opposite view on Tesla's AutoPilot branded self-driving feature. We think Tesla's urge to make AutoPilot magical is unnecessary because Tesla has been basically pretty groundbreaking regardless of whatever it's been doing on the self-driving stuff.

                                                            The AutoPilot functionality, again that's what Tesla calls its autonomy future isn't really that differentiated. It was operational a good year after vehicles with steering assists were introduced by Mercedes. I know, I had one of these Mercedes, a 2015 E-Class.

                                                            I remember I swapped it, I swapped my car with the Tesla owning friend in about 2016 and we both thought the Mercedes had the better autonomy features. This fact of course is something you cannot bring up to a Tesla owner or pretty much anywhere in the Bay Area and have it be, remotely have a chance of getting accepted in light of what the sort of the religious view that people have on Tesla's autonomy features.

                                                            You'll notice that the company has been called out quite a fair bit recently, and by that I mean mostly in 2018, regarding accidents where AutoPilot seemed to have had some role to play. 

                                                            Two things we often hear in Tesla's defense sometimes delivered by Tesla itself go something like the following. The first offense tends to be that the driver wasn't using the system correctly when an accident happened without a pilot engaged. 

                                                            For example, that he, and it always is a he, was getting warnings from the system, didn't read the manual about not leaving the system unsupervised, and so on, but the protestations of Tesla's competition around the AutoPilot issue are interesting to take note here.

                                                            They seriously question the way Tesla has evangelized the technology and they actually view the very phrase, AutoPilot, as giving the driver a sense of complacency and contributing to some of these accidents.

                                                            For example, AutoPilot was introduced with fewer safeguards than that Mercedes system I mentioned. The Mercedes system from the very beginning demanded regular driver input mainly in the form of the driver introducing some torque steer into the steering wheel about every 10 seconds.

                                                            Tesla didn't have this safe guard in it. The company later backtracked, we've got similar safeguards, but to some extent the damage had already been done in terms of what the customer had been programmed to think was the capability of their vehicle.

                                                            Take a look at on YouTube and it's almost not believable now, but you'll find early videos of drivers playing patty-cakes while their Tesla is, "Driving itself," right after the company released their initial version of AutoPilot software. 

                                                            One of the first fatal AutoPilot mode accidents was a driver who had been watching a movie on his iPad at the time of the accident. If you look at how other manufacturers position similar systems there is little chance that any other makes' drivers would feel comfortable doing anything of the sort.

                                                            What seems to have happened by the way in that accident when a truck laterally moved into Tesla's path and the car, the Tesla drove itself basically into the undercarriage of that truck seems to be the following.

                                                            In look ahead mode a car is typically integrating input from the camera which is responsible for sort of the upper half of the vision of the vehicle and integrate that input with their radar which has responsibility for the lower half of the forward vision.

                                                            I'm exaggerating a little bit, but that's roughly how it breaks out. It seems like the camera had looked at the washed out sky in a bright sunlight and equated it or at least couldn't tell the difference between that sky and the white side board of the truck, it all look like sky at the top end.

                                                            The radar recently has been tuned to look fairly directly ahead as opposed to more to the sides. This is true of all auto manufacturers, they were getting too much noise and backscatter when they had both wide angle of view and high angle of view from their radars, therefore the camera didn't feel there was a problem until the car had already moved into the truck ... excuse me, had already moved into the driving line of the Tesla and by then it was too late for the radar to capture the truck and issue an emergency stop.

                                                            A second line of defense is and this one typically comes from Tesla itself, that despite the accidents there's a net benefit anyway. Here's a quote from the company, " For Tesla there is one fatality including known pedestrian fatalities every 320 million miles in vehicles equipped with AutoPilot hardware. If you are driving a Tesla equipped with AutoPilot hardware you are 3.7 times less likely to be involved in a fatal accident."

                                                            This comparison really is not a good one at all if you know the details of the numbers. Their overall math is correct, Tesla’s one in every 320 million miles between fatal accidents compares favorably to about one in every 100 million, 80 million, we'll just round it up to 100 million miles on average for vehicles overall, but one in 100 million miles ... excuse me, one fatality every 100 million miles is an average.

                                                            If you're in a vehicle that's heavier or a sedan typically driven by a middle-aged person or a sedan typically ... or any car really driven by a wealthier person the track record for those vehicles is much better than one in every 100 million miles. Years go by before there's a fatality in vehicles of this class.

                                                            For example fatalities for example regarding the heavier vehicles your one third less likely to get killed in an SUV, on our middle-aged point 40% of fatalities are attributable to people below 25 or over 70 which probably comprise much less than 40% of the driver community.

                                                            Years go by when there's a death in vehicles like this. Forbes recently had an article that showed for the years 2012 through 2015 the following vehicles amongst others had no fatalities in the US. The Audi midsize sedan, the A6, their small SUV, the Q5, a couple Lexus models, midsized Mercedes SUV and a Jeep.

                                                            Obviously by the way there is a lot more Tesla’s running around with AutoPilot hardware than Tesla’s actually driving in AutoPilot mode. And so, when you make these adjustments that we've been talking about and whatever factor you want to apply to it, the math looks a lot less favorable and even worrisome for Tesla owners driving in AutoPilot mode.

                                                            We would be remiss of course if we didn't point out where there was some potential for spectacular achievement by Tesla in their particular approach to self-driving. Those are the quotes we gave you just a minute ago from Tesla it said, "In vehicles equipped with AutoPilot hardware." 

                                                            What's interesting with Tesla is they include this AutoPilot self-driving hardware regardless of whether the option is selected and paid for by a customer. Therefore, all these vehicles whether you've paid for AutoPilot or not are collecting your driving data on a scale that no other automobile firm can match.

                                                            Tesla's data collection is an enormous advantage. This is because the self-driving autonomous systems are desperate for data on how you are driving. These system don't follow a set of algorithms, they're not programed to be able to drive per se, but rather they improve their behavior over time by comparing their own guesses, which are initially pretty simplistic of what the car should do to what a human actually does in the same situation.

                                                            If Tesla's data advantage allows it to achieve high levels of autonomy it'll be a really spectacular achievement, not only because it's achieved that high level of autonomy, kind of like Waymo and maybe GM Cruise are achieving now, but because it would have done so without the inclusion of an expensive, kind of pesky technology called lidar.

                                                            All other OEMs are betting and assuming that this rather expensive lidar technology, which is not nearly close to being available on a consumer scale, they're all assuming that it's going to be critical for advanced autonomy.

                                                            It would be quite end-around for Tesla to get self- driving or significant self-driving modality done without the inclusion of that, now essentially a bottleneck lidar hardware.

                                                            Tesla's ability by the way to include this hardware without you paying for it kind of puts its competition in a jam. The competition is just too much insensitive to include hardware for which they're not paid just to collect data, that's because their public investors don't give them the leeway that Tesla has in terms of being loss-making and essentially investing the way that the Tesla does.

                                                            GM just can't throw in 10 cameras, which is what the Model 3 has unless it's been paid for it  either as an option or through a better sales price because it's an important feature to their particular buyer group. Neither of these is a problem for Tesla. 

                                                            Now one workaround that the financial markets are trying to bring to the table is for the venture markets, which are more forgiving financial markets, to subsidize hardware deployment.

                                                            For example SoftBank, the major Japanese investment group just put in about 150 million into an auto vision startup called Nauto mainly so that that company could give away its hardware to the car companies like GM and others so that they could get installed in the vehicle, it wouldn't be quite the out-of-pocket cost, if you will, for GM as it otherwise would be, and the two companies GM and Nauto then share the benefits of that data collection.

                                                            Okay ladies and gentlemen I think we’re going to end it here. Hopefully we've provided you a lot of things to provoke your thinking and hopefully we've provided you a foundation to reach your own conclusions specific to your own interests about the world of electric vehicles.

                                                            You know an inflection point in any one of the things we've talked about in our lifetime would be amazing. The change in the way that cars are powered, combustion to electric, that's a once in a lifetime change. The change in the way they're produced, all the changes to the supply chain we talked about, that would be an interesting once in a lifetime change.

                                                            A disruption of the dealership model, same thing. A disruption in how cars are driven, their ability to drive themselves, that would be a pretty interesting once in a lifetime change, that all are either happening or being attempted concurrently is astonishing, and that a single company, and one of the newest, Tesla, has a role in all of them, it is almost not believable.