This story was originally published in Worth.
I’ve written recently about how our brains are by nature set up for linear thinking, and how that is something that takes work and persistence to overcome. Meanwhile, though, this property of human cognition creates market inefficiencies that can be exploited, and Tesla provides the ideal case study. The executive summary for innovation-facing investors is this: Owning Tesla acts like a call option on a lot of emerging and future tech.
Let’s consider Tesla’s lead in autonomous driving. The type of AI Tesla is employing to get us from A to B while we’re napping is a method known as reinforcement learning. This is the same approach used by Google Deepmind’s AlphaGo Zero which, according to Singularity Hub, “allows machines to accumulate thousands of years of knowledge in a matter of hours.” This type of AI pattern recognition stands in contrast to the brute force method used by most autonomous driving competitors of simply referring to a detailed map of the world.
To deploy reinforcement learning, Tesla has developed the world’s most advanced AI-driving dedicated processor, which is itself another innovation and piece of IP that Tesla can and will capitalize on. In its recent Autonomy Day event, Tesla claimed its “Full Self-Driving” chip, capable of running at 144 trillion operations per second, is significantly more powerful than Nvidia’s current top-performer, the Drive Xavier chip, which runs at 21 trillion operations per second. In addition, the Tesla system has a huge advantage in terms of real-world data collected, with 425,000 vehicles constantly capturing inputs to learn from and more Teslas hitting the road every day. Is there value in self-driving technology? Surely there is, but where the self-driving unit of Alphabet, Waymo, is independently valued as high as $250 billion, I don’t see much evidence that Tesla’s huge advantages in this field are being valued at all.
In autonomous driving, the neural-net human learning approach will be far superior to the raw brute force database of world map approach, and those 425,000 cars and counting are providing a lot of data for the neural net to learn from. But are we worried about United States regulators never approving autonomy? Well, then what about the larger market of China? What about the UAE and Dubai? And what about the clear, economically advantageous autonomous driving applications for Tesla’s upcoming semi-truck? You don’t need to rely on any one use case or particular market to think Tesla’s technology will be monetizable.
Tesla is also leading in vehicle efficiency, getting far more range out of each watt than any other company making an electric car. This means they have tech advantage, part of which is in the drivetrain, and part in the electric motors themselves. Noted engineer and auto industry consultant Sandy Munro broke down some of the tech for the Autoline network in a January interview. Munro particularly notes on the superiority of Tesla’s permanent magnet motors: “This is a science all by itself…how come this thing’s lighter and costs less, and it has more power—what the hell?” Munro would like to understand better how it works, but “the thing is, I can’t reverse engineer this damn thing!”
Tesla’s battery pack is also the most energy dense in the industry, by cost and by weight. At Autonomy Day, Tesla mentioned that its upcoming version of its model S and Model 3 battery packs will be good for one million miles of driving before they need to be recycled (company tests also show their electric motors lasting the equivalent of a million miles). As a result of these techs, Tesla cars travel farther on every watt of battery-stored energy, delivering far more range for each equivalent battery size. The old 2012 Tesla Model S gets more range out of a smaller battery pack than either the 2018 Audi E-Tron or 2018 Jaguar i-Pace, implying Tesla has a minimum six-year lead in battery and drivetrain efficiency tech. The 2018 Teslas have more range still.
As well, the total addressable market of these advanced batteries goes beyond cars and trucks, to home and utility-scale power storage, both of which are booming businesses.
Other examples of Tesla innovation are perhaps less sexy, but still important, and even game changing, such as the Superbottle, which is the secret sauce in Tesla’s industry-leading integrated thermal battery management and HVAC systems. Automotive news website Jalopnik describes the Superbottle as “different in that it packages two pumps, one heat exchanger and a coolant control valve all within the bottle itself…incorporating a valve, a heat exchanger and coolant pumps all into one device.” All that means easier serviceability, weight savings, fewer parts, reduced assembly costs and time, and therefore cost savings.
Vehicle longevity, minimal maintenance and safer driving with assisted autonomy also mean that Tesla has the opportunity to innovate in the high-margin insurance business.
So Tesla has a large lead in automotive hardware, software, battery production and technology, motor design and efficiency, and in any number of other areas like Superbottle. And it delivers all these advantages in each vehicle. Paraphrasing something Elon Musk said on the last earnings call, it is amazing to me that Tesla released the Model S in 2012, and seven years later it’s still waiting for that car to have a meaningful competitor.
And let’s not forget about SpaceX, which is another innovation powerhouse, and one with a free interflow of ideas and technologies with Tesla. The unforgiving environment of outer space has historically been the mother of some of our most useful innovations, including CAT scans, LEDs, wireless headsets, notebook computers, memory foam and camera phones (the list goes on). With ambitions far beyond the present paying gig of resupplying the International Space Station, SpaceX almost cannot help but continue in this grand tradition of dramatic tech evolution.
I’m not going to dive into Tesla’s solar business in this post, but it is important to recall that they are America’s third largest home installer, and are leading innovators here as well, with the solar roof product potentially providing power while allaying HOA curb-appeal concerns in the near future.
Clearly, Tesla is much more than a car company. It is a core-deep tech company. But with our attention spans now shorter than ever, and with the world’s critical issues and the data smog around them more complex, we are more susceptible than ever to misinformation and to our natural linear thinking. We have a hard time appreciating that parabolic change is upon us, and one result is that many analysts, pundits and Wall Street mainstays are failing to correctly value the innovation that Tesla represents. But this presents opportunity: misinformation and rapid change make for inefficient markets. Modern investment processes should bear in mind the directional arrow history is taking.
Garvin Jabusch is the co-founder and chief investment officer of Boulder, Colo.-based Green Alpha Advisors.