March 24th, 2019
This is mostly a reaction piece to Benedict Evans’ piece Tesla, software and disruption.
I am a big Tesla fan, and believe that electric cars and automation are the future. However, I am no expert on any of these issues which is why I found Evans’ article so interesting.
First, he breaks down Tesla’s interesting characteristics into four main categories: batteries and motors, software and integration, Tesla’s ‘experience’ disruption, and autonomy. Characteristics two and four interested me the most.
To me, Tesla’s ability to send software updates to cars overnight has been underappreciated by the general population. Evans makes a point that ‘we will go from complex cars with simple software to simple cars with complex software’. This seems to me to be a reasonable next step, but I can see other car companies struggling with the transition. Furthermore, I agree with him that it is likely that traditional car companies will trivialize the difficulty of integrating separate systems. It took a week, a couple thousand dollars, and a loaner car to see why my family’s Infinity FX 45’s ignition seems to stick when turning on sometimes. Tesla improved the braking distance of the Model 3 by 19 feet overnight.
This change is transformational both practically and socially. Technology consistently pushes the status quo forward, and I can think of very few instances where society has chosen the more primitive technology (looking at you, coal vs nuclear energy). While there may be parallels between coal lobbying and the car industry, I believe that the shift to software updates for cars is inevitable. If so, Tesla’s engineering lead could lead to some fascinating outcomes.
Potential outcome one: traditional car companies attempt to emulate Tesla’s software and bungle it. Evans notes that those car companies may have to go through a ‘cycle of learning… before buying it from someone who does it better’. Evans notes astutely how Apple and Tesla are very specific with optimizing components working in conjunction, while Windows computers and traditional car companies use more interchangeable parts. From blue screens to unexpected shutdowns, most people have had at least one frustrating experience with Windows updates. While that is inconvenient for a computer, what happens if it happens to your car? Nike’s self-lacing shoes bricked after faulty firmware, completely removing its main functionality. A pair of shoes, or to a certain extent a computer, can be swapped out. For most people, however, they cannot swap out cars. This would make faulty firmware a significant problem for traditional car companies.
Finally, autonomy. When it comes to cracking the code for driverless cars, consensus says the more data the better. For now, Waymo seems to be in the lead for driverless cars, but Tesla has logged 100x more miles of autonomous driving. But I think there’s an even more important issue than miles logged.
I think it’s likely that we will see see the driverless revolution result in an oligopoly, with many traditional locomotive companies driven out of business. First of all, cabs. Waymo is continually expanding it’s taxi service, and Uber’s dream is to have a driverless workforce. Just like how New York taxi cabs are separated into yellow and green cabs, I foresee a future with Waymos and Cruises (insert your own favorite driverless car companies here, but the point still stands).
What does this have to do with Tesla? I think the consumer car industry will go through a similar revolution. However, there are additional barriers of entry. Not only will the consumer autonomous car need to have safe, functioning software (and all the challenges that come with what I mentioned above), but they will need to be produced on a large scale. I believe this is where Tesla has a large competitive advantage. They not only have the data from their fleet of autonomous vehicles on the road currently and the software capability/practice born from all their models, but they also have the capability to produce cars on a mass scale. Granted, they are still struggling to produce fast enough with quality, but their gigafactory and factory in Fremont give them a production advantage over Waymo’s plant in Michigan.
This leaves them in an interesting position. I have little confidence that all traditional car manufacturers will be able to make the transition to electric cars, much less the autonomous world. On the flip side, Alphabet’s strength has never been in their consumer-facing hardware. If they create software for autonomous driving first, I could see them diverting significant resources to devloping their supply chain and production. After all, it can be hard to stop a trillion dollar company. But even then, I think there is room for Tesla to play. Ford and GM? I’m not so sure.
Tesla is not without its problems. If these lawsuits serve to depress their stock price for now, that’s better for me. I believe, however, in the long term vision of Tesla and their products. Maybe I’m the fool for believing and buying while they crash, or maybe I’m just following the age old advice of buying underpriced value. Time will tell.