Musks Self-Driving Taxi Needs a Time Machine

Musks Self-Driving Taxi Needs a Time MachineOn October 8th, news broke that Tesla will hold a "self-driving taxi launch event" in Los Angeles on October 10th. CEO Elon Musk needs to wow his audience to mask the fact that his self-driving technology has yet to materialize

Musks Self-Driving Taxi Needs a Time Machine

On October 8th, news broke that Tesla will hold a "self-driving taxi launch event" in Los Angeles on October 10th. CEO Elon Musk needs to wow his audience to mask the fact that his self-driving technology has yet to materialize. Musk first laid out the roadmap for self-driving during the company's "Autonomy Day" in March 2019. Now, at the event this Thursday, he needs to keep the story going.

Five years ago, Musk envisioned a self-driving taxi, a marvel that would drive itself, generate revenue, and make money for its owner. He claimed the vehicle would be a reality by the end of 2020. Even accounting for the pandemic, that goal has been severely delayed.

During this period, Tesla's electric vehicle sales, while growing fivefold at one point, have since leveled off. This is reflected in the company's stock valuation, which climbed from less than $50 billion to a peak of $1.2 trillion before settling back down to about $400 billion. As the electric vehicle market cooled, artificial intelligence filled the gap. Bullish analysts and Musk himself often attribute a small portion of Tesla's value to electric vehicle sales, with the majority tied to dreams of self-driving taxis and robots.

Musk declared October 10th "a day for the history books" and even used the slogan "We, the Robots." It's worth noting that the event stems from a social media squabble on X. Tesla held an Investor Day in March 2023, showcasing Musk's "Master Plan Part 3," which emphasized developing a cheaper electric vehicle. However, about a year later, Reuters reported that the cheaper electric vehicle project was canceled. At the time, Tesla's stock was struggling due to the company abandoning its electric vehicle growth goals, and the Reuters report further dampened market confidence. Musk accused Reuters of lying, but then abruptly announced an "autonomous taxi launch" on August 8th, indirectly confirming the report and introducing a new Master Plan. The launch was ultimately pushed back two months, fueling further doubts about Teslas execution. If Musk announces yet another new plan this week, the lifespan of his "Master Plan" will become extremely short.

Tesla's humanoid robot, Optimus, is expected to take center stage as a future highlight. However, nobody expects Tesla to announce they have solved the problem they set out to solve five years ago: developing a self-driving taxi driven by general artificial intelligence that Tesla is willing to be responsible for. This vehicle relies on cameras and conventional radar, not expensive lidar technology, and boasts wireless software updates that could enable millions of Teslas across the US to drive autonomously.

Even Adam Jonas, a Morgan Stanley analyst who is highly bullish on Tesla, doesnt hold out that kind of hope. In his latest report, he predicts that the event might showcase a connected taxi capable of self-driving on closed Tesla-defined routes. He also expects Tesla to introduce a supervised FSD ride-hailing service, where Tesla owners with the Full Self-Driving (FSD) package offer Uber-like services but allow the car to drive itself in certain situations. This way, Tesla could collect more self-driving data, but its hard to imagine passengers choosing this service unless it comes at a steep discount.

All of these signs indicate a retreat for Tesla in the self-driving space, a repositioning of the company in a fast-changing market that also impacts the current leader, Waymo.

Waymo has been steadily expanding its business, albeit in a much smarter way for some time. The company recently announced that it is expanding its self-driving taxi service to Austin and Atlanta, building on its Phoenix base and partnering with Uber. This allows Waymo to tap into Uber's vast customer base and scale its fleet based on demand, maximizing utilization. Meanwhile, Uber's flexible driver workforce can handle peak demand and lower Waymos unit costs. For Uber, the partnership offers a new differentiated vehicle and makes them a reliable partner in the space after their own self-driving efforts faltered.

In a sense, Waymo is more like the tortoise in this race, deploying a small number of self-driving taxis in limited geographies, relying on remote operators for support, and gradually building up autonomous capabilities. While slow, this approach offers higher safety. Self-driving expert Alex Roy, founder of risk fund New Industry Management, favors Waymo but points out that if Tesla ultimately achieves general self-driving through lightweight hardware, Waymo would be in survival mode.

Achieving the general self-driving Tesla claims is still a difficult and expensive task. Philip Koopman, an associate professor at Carnegie Mellon University and expert in autonomous driving, argues that the automotive industry has underestimated the complexity of all the potential unexpected events on the road. For true safety, many situations need to be addressed, and these situations rarely occur for machines. Its not about how rare something is, but how likely it is to happen to you, Koopman summarizes. Therefore, Tesla clearly needs more time.

Roy recently wrote an article about Tesla's "narrative control" in which he describes how Tesla has redefined societys view of electric vehicles and autonomous driving through substantial progress in battery technology and vehicle design, while simultaneously popularizing terms such as autonomous driving, robo-taxis, and superchargers. Despite missing many stated goals, the company has maintained this ability to bend reality. Crucially, the stock has continued to grow in spectacular fashion even as the companys core business has stagnated. This is the position the company hopes to solidify with its upcoming Los Angeles celebration, in the face of undeniable progress by its competitors.

One way to continue doing that might be to release data showing that Teslas driver assistance features are significantly safer than humans, or by revealing how many Tesla buyers actually pay extra for FSD software. Oddly, a company that cites safe autonomous driving as its primary investment rationale is hesitant to back it up with a few simple data points. Perhaps Musk will still prioritize spectacle over substance at the event on Thursday. As always, the question is whether investors will be fooled again, forgetting about the broken promises.