Tesla’s AI Gamble: Analyzing the ‘We, Robot’ Event

By Garvin Jabusch.

On October 10th, Tesla hosted its highly anticipated “We, Robot” event on the Warner Brothers lot in Hollywood. The glitzy showcase offered a glimpse into Tesla’s vision of a future of autonomous transportation and artificial intelligence. I showed up to see and hear firsthand the revelations and promises made by Elon Musk and company. In this post, I’ll break down the key announcements, analyze their potential impact, and offer my thoughts on Tesla’s ambitions.

Cybercab Interior. Photo: Garvin Jabusch

The Cybercab

The Cybercab subjectively to me looks cool: small, sleek, simple, aerodynamic. No-nonsense transportation but with a sense of style. It’s a two-seater with butterfly doors, a reasonably spacious trunk, and, as advertised, no wheel, pedals, or controls other than the screen.

Lack of Technical Details

But details about what’s under the skin and other specs were sparse. Addressing the group, Elon Musk gave a cost per mile figure of $0.20, said the vehicle would cost under $30,000, and noted that it would include Tesla FSD hardware version 5 (current Tesla’s are shipping with hardware 4). And… that was it in terms of specs. They’re not even citing estimated range.

Musk’s slide deck showed off an inductive charging system, which would allow the Cyber cab to merely park in a spot and be charged, no human intervention required, and a robotic cleaning system, which looked like it would work well with the butterfly doors open.

Tesla gave us a chance to experience the Cybercab, and in the demonstration ride itself, the vehicle seemed to work perfectly fine. But it was on a closed, highly controlled course of maybe half a mile, and we never exceeded 5 mph. The experience wasn’t too dissimilar from my memory of going on the “It’s a Small World” ride at Disneyland when I was 10 years old – in 1977. The lack of wheel and pedals did not seem especially revolutionary in that context.

The Challenge of Full Autonomy

But this brings up another important point: No wheel or pedals means Tesla must actually solve completely autonomous robotic driving before they can sell even one of these things. Musk said from the stage that they will have finally cracked that problem by the end of next year. But it’s hard to take him too seriously, because as we all know, he’s been promising that since 2016. Tesla’s AI and super computing infrastructure and capabilities continue to get better, so at some point they may actually solve truly autonomous driving, but I would hate to be the oddsmaker setting the over/under on when that will be. In any case, Elon said he thinks California and Texas will be the first places they offer the Robotaxi service, and he said it will be by the end of next year, pretty much for sure. He also asked us to believe that Tesla’s entire fleet, including existing cars, would achieve full autonomy before Robotaxi’s projected 2026-2027 launch. Draw your own conclusions about full level 5 autonomous driving (not geofenced a la Waymo), and regulatory approval (e.g., producing a steering wheel-less Cybercab will require waivers from the federal government) both occurring by the end of next year. For reference, Elon clarified onstage that for his purposes, level 5 means being able to get in your car, set the destination, fall asleep, and wake up when you’ve arrived.

AEI official driving automation nomenclature. Image: https://www.sae.org/blog/sae-j3016-update 

Company Risks and Strategy

In terms of company risk, what makes these promises different from the many promises of full self-driving Musk has made in the past is that this time he has dedicated real resources to a production line for making the Cyber Cab. Deploying the capital expenditure toward building a car that entirely depends on a so far non-existent technology strikes me as a heck of a dice roll. I mean, why not, as an intermediate step, release a geofenced version just so you’re not staking so much on a fully autonomous version that may still be years away? Tesla’s huge new data center is coming online in Texas next year, and maybe that will enable a large leap forward in full autonomy, but that’s the problem, we really don’t know.

To defray this risk, Tesla should simply make a version with a steering wheel, because the existence of a cyber cab (the first 20 of which were present at the event) with a <$30,000 price tag shows they may know how to manufacture that. Why not make and sell a drivable version to defray some of this risk? True, perhaps introducing a specialized Robotaxi model, dedicated solely to autonomous ride-hailing, would prove more cost-effective and less complex than adapting the vehicle design for a human driver. Okay, but only if Level 5 works relatively soon.

It seems to me that a solid, bordering on cool, EV under $30k will have a large potential market in a world where cheap yet great Chinese EVs are popular enough to inspire draconian protectionist trade measures in the US and Europe. It’s simple: people really just do want a cheap EV almost anyone can buy.

Robovan concept. Photo: Garvin Jabusch

The Robovan Unveiling

The event also included the unveiling of a new Robovan (which Musk pronounced Ruh-bow-vun), but here again there was very little to nothing in the way of specs or operating information, not even a potential timeline for production. It is Art Deco-inspired, and subjectively, to me, does look kind of cool. But as one equity analyst who was also there covering the event said to me, “sure, it looks cool. If they ever make it!” Also note that there are no controls within the van, meaning level 5 autonomy has to be cracked for the van to be a thing, as well.

Tesla has not made reservations possible for cyber cabs or vans, and it seems like neither product is really that close to production. Which begs the question, why did they have the reveal event now? Why not wait until things are more fully baked, and actually open the order book? Unlike with previous launch events, there’s no pre-order page for interested parties to visit.

Optimus Humanoid Robots

The Optimus humanoid robots also made an appearance and seemed to work okay doing things like dancing and serving drinks. Discussion at the event was that they were being remotely controlled by humans, so maybe, like with full self-driving as it is today, humanoid robot autonomy kind of works but still requires human assistance. In any case, Tesla was not clear about that, so we don’t know what Optimus is capable of today. So, there’s no telling if and when Optimus will be a real, useful product you can buy (Musk quoted $20-to-$30,000), but if it does become one, it could be a legitimately huge source of revenue for Tesla. Which you could think of, variously, as an exciting shot on goal for potential future revenue growth, or, as yet another bunch of capex and brain power being spent on something that hinges on the critical dependency of real-world embodied AI being able to navigate better than a human can.

Musk’s Vision of the Future

An overarching point Musk was making though was about the inefficiency and unsustainability of our current transportation systems and human labor capacity. Musk was asking us to imagine a world where our commutes are free from traffic jams and parking woes, where self-driving vehicles reduce accidents and fatalities, and where cities can reclaim space for parks, businesses, and homes. He invited us to consider that the long-term goal of Optimus is to give humans back valuable time, whether through help around the house or assistance in factories.

Bottom Line

Tesla’s “We, Robot” event painted a picture of a future where autonomous transportation and artificial intelligence seamlessly integrate into our daily lives. The vision of traffic-free commutes, safer roads, and reclaimed urban spaces, and freed up human time and labor, is undeniably appealing. However, the lack of concrete details, optimistic timelines, and the company’s history of unfulfilled promises in the realm of full self-driving technology give me significant pause. While Tesla’s ambitious projects like the Cybercab, Robovan, and Optimus robot do show the company’s still-innovative spirit, the path from concept to reality remains fraught with technological, regulatory, and practical challenges. The road to an AI-driven world is longer than Elon Musk and his team suggest.

###

Green Alpha is a registered trademark of Green Alpha Advisors, LLC. Green Alpha Investments is a registered trade name of Green Alpha Advisors, LLC. Green Alpha also owns the trademarks to “Next Economy,” “Investing in the Next Economy,” “Investing for the Next Economy,” “Next Economy Portfolio Theory,” and “Next Economics.” Green Alpha Advisors, LLC is an investment advisor registered with the U.S. SEC Registration as an investment advisor does not imply any certain level of skill or training. Nothing in this post should be construed to be individual investment, tax, or other personalized financial advice. Please see additional important disclosures here: https://greenalphaadvisors.com/about-us/legal-disclaimers/

At the time this article was written and published, some Green Alpha client portfolios held long positions in Tesla (ticker TSLA). These do not represent all of the securities purchased, sold, or recommended for advisory clients. You may request a list of all recommendations made by Green Alpha in the past year by emailing a request to any of us. It should not be assumed that the recommendations made in the past or future were or will be profitable or will equal the performance of the securities cited as examples in this article. Not all Green Alpha separate accounts or our sub-advised mutual fund held the stock mentioned. To inquire whether a specific Green Alpha portfolio(s) holds stock in any particular company, please call or email us.