Tesla vs. world on self driving tech

Silicon Valley.

My Tesla, sitting in my garage at the moment, costs about $5 an hour (actually it’s closer to $2 but I was being generous and keeping the math easier to get). If you bought one you’ll pay about $1,000 a month in car payments. About $200 in car insurance. About $100 (or far less if you drive less than I do, since I put 31,000 miles on mine in its first year) for electricity. About $1,200 for tires every 20,000 miles or so (hey, it is a lot more fun to drive than my Prius was). And a few extra bucks for maintainance (it still needs some, even though you need far less brakes due to using the electric motor for most braking and no oil changes). Divide that all up by the hours in a month and you come to less than $5 an hour.

Now, what does an Uber cost? Pretty close to the same, but you need a driver which costs about $15 an hour in Silicon Valley. So, let’s say that’s $20 an hour.

That shows you how big a threat that is to Uber/Lyft’s business model.

But what if you could send your car to pick up your laundry? 
Or pick up the stuff you just ordered from Amazon?
Or pick up dinner for you or your family?

What if you could send your car to do work on your behalf? It would start paying you back.

I don’t believe Elon will get there next year. The dude is always too aggressive on timing.

But it is coming. Is it three years away? Five? Certainly no more than 10.

Why can’t the other car companies do it? They don’t have a fleet on the road with an AI chip yet. A neural network inside. An exponential learning system.

They also insist on using knobs on their dash and refuse to force its owners to use their phones as keys. Both of which will pay HUGE dividends when the Tesla network actually starts fully self driving. They say their customers won’t allow such things. Even my dad turns off his phone, so I see their point. Yet my Tesla already has both, which means everything on the car can be customized to you simply by carrying a phone around.

One other thing I’ve been doing? Research on how people approach robotaxis. When I ask them:

“Are you willing to get into a car without a steering wheel?” (Like Waymo or Zoox or any of the others that are doing research will force you to accept). The answers are almost wholly negative. “F**k no,” many answer. Many explain they don’t even like using cruise control and that they actually like to drive.

Then I ask a second question: “What if the car drove to you and then you had the choice to drive it away?”

The answers always change to “yes, I’m cool with that.”

Now, I know that they all are lying. Why? Peter Norvig, Google’s head of R&D told me he had the data. He knows that everyone will stop driving their cars after they have about three days of experience inside a fully-self driving car.

He asked me a question while we had dinner a few years back that provides the key.

“Do you know why Google made their own small cars that could only go 25 miles an hour instead of giving everyone big cars with all the LIDARS and that?”

“No, why?”

“Because when we gave the full size cars to employees we had them sign a contract that said they would be fired if they took their eyes off the road or their hands off the steering wheel. We showed them the sensors and cameras.”

“They all broke the rules after three days.”

“Despite the threats of being fired? Wow.”


So, this is a huge marketing problem. Waymo and other self-driving car companies will need to do a lot of marketing to convince people to use it.

Tesla won’t.

I remember sitting next to Kraft food execs and they say they spend $34 per person to acquire a customer. Tesla doesn’t spend much at all in marketing, if any. So, Tesla will be more profitable out the gate than, say, Waymo will be.

And, even better, everyone knows that a Tesla is a HOOT to drive (everyone who gets into mine says “wow” because it accelerates so fast and mine is one of the slower ones in the fleet. Some Teslas go 0-60 in 1.8 seconds, mine does it in 5.3).

All this translates to: when Tesla turns on the self driving network it will do real damage to Uber and Lyft and I don’t see how Uber and Lyft will be able to quickly react since Uber and Lyft don’t make their own cars, don’t own their own battery plant, and if they try to work with Toyota, Ford, Mercedes, VW, et al, they will find slow-moving car companies that resist making the tech and other changes that they will need to do to compete with Tesla.

That said, it’s not too late yet. My friend Brandon Wirtz says that the chip architecture that Tesla chose is the wrong one and that the bet against LIDAR is the wrong bet. If he’s right (and he usually is) that gives the others time to fix their internal cultures and get their own designs on the road.

But what if he’s wrong? Tesla will end up with a transportation monopoly that will be chewing through Lyft and Uber, and a Tesla pickup truck is coming in a few years.

We visited Ford’s plant last summer where they make F150 trucks. That makes a new truck every minute. Now, what happens if Tesla takes 25% marketshare, like I expect it to do with an electric truck?

Huge disruptions. (That’s Silicon Valley’s word for job losses).

And right around that same time Tesla will be shipping 18-wheel trucks too and disrupting shipping.

A perfect storm for the transportation industry is right around the corner.

We could see a major change in America that only a few are warning about and it’ll all happen in the next decade.

Today’s chip just shows the kind of effort that is being done to get there. I know Tesla isn’t alone in investing billions in this stuff. But it sure is the most exciting company on the field today because it is figuring out the pieces fastest.

I wonder what self-driving expert Brad Templeton thinks. Oh, he just posted he was impressed by Tesla’s announcements today.