I’m a hobbyist investor with:
A high risk tolerance (i.e. a willingness to lose money)
A long time horizon (i.e. I’m fine holding a stock for 5 years or more through ups and downs)
A strong personal conviction in the potential of deep learning for robotics
As far as publicly traded companies go, self-driving cars are one of the few opportunities — perhaps the only opportunity — for retail investors like myself to place a bet on deep learning of robotics. At least, as far as I’m aware today.
I’ve toyed with the idea of a portfolio that looks something like this:
15% Alphabet (Waymo)
15% General Motors (Cruise)
15% Baidu (Apollo)
15% Intel (Mobileye)
15% Nvidia (chip supplier for automakers)
ARK Invest also has an actively managed ETF that focuses on autonomous vehicles, among other areas. Currently, the fund’s holdings include (in order from larger to smaller positions) Tesla, Aptiv, Baidu, Nvidia, Alphabet, and General Motors.
Right now, my portfolio is 100% Tesla based on the view that Tesla has a data advantage that extends through perception, prediction, and planning. Tesla has the largest deployment of four-wheeled robots — perhaps any genus of robots — in the world. Oddly, this seems to have been largely overlooked by analysts, investors, and the media.
Recently, I had the small revelation that lidar is a bit of red herring. If Tesla hits a wall with its non-lidar approach, it can always just start using lidar in a small test fleet like Waymo and Cruise. The data advantage will still be there for all the non-lidar aspects of the system.
If Waymo or someone else gets to commercially viable robotaxis before Tesla by using lidar, Tesla can start using lidar and be a fast follower. That data advantage will help it be fast.
Overall, I don’t see competition as the main risk. I see feasibility of the technology as by far the main source of risk and uncertainty. So, when I see encouraging signs from any company in this space, I feel good. A rising tide lifts all boats. Or at least the expected value of the boats, considered in the aggregate.
ARK Invest has an interesting set of scenarios for Tesla:
ARK gives Tesla a 30% chance of launching a robotaxi service and a 70% chance of remaining a simple carmaker. But in the robotaxi scenarios, Tesla is worth $15,000 a share. Specific probabilities and price targets aside, I feel like this is a healthy way to think about the opportunity. A relatively small but real chance of an outlandishly large outcome. I’m personally comfortable with that sort of gamble. Not everyone is. And that's okay.
What about other investment opportunities besides self-driving cars? Where is deep learning being applied to robots? Covariant is doing warehouse logistics, but it’s a private startup. If I want to place a bet on that opportunity, what companies should I look into? Amazon? If anyone has an idea, please let me know.
What I’m thinking about lately:
The coming self-supervised revolution in computer vision
How human interventions can train neural networks to understand causality
What internal metrics give autonomous vehicle CEOs such confidence that true self-driving is imminent
If you'd like me to write an email on any of these topics, let me know and I’d be happy to oblige.
Financial disclosure: I own shares of Tesla (TSLA).
Important disclaimer: This email is not intended as financial advice. Invest at your own risk and please consult a professional investment advisor if that is appropriate to your situation.