Tesla Is In Trouble. Again.
Tesla Is In Trouble. Again.
A core pillar of its future has just been undermined.
Tesla is in a spot of bother, and I’m not talking about the fact they just announced they are laying off more than 10% of their global workforce. That is a surface blemish compared to the core-shattering headache Tesla has just stumbled into. You see, not only are their sales dropping faster than Musk’s reputation, but they have slashed the price of Full Self-Driving (FSD) from $199 a month to $99 a month to get more income through the door to make up for their terrible sales. This is a minor detail to the casual observer, but to anyone who has been paying attention to Musk’s antics over the years, this is more like Tesla shooting itself in the foot with a nuclear bazooka. Let me explain.
Musk has stated many times before that without self-driving cars, Tesla would be worth basically nothing. While that is undoubtedly hyperbole, it’s not that far from the truth. You see, FSD, which is meant to improve over time into a fully fledged self-driving car, used to cost customers $199 per month or a one-off payment of $12,000. However, Elon thinks that FSD is massively underpriced and claims its actual value is somewhere over $100,000 per car. Investors have bought into Musk’s predictions, and Cathy Woods of Ark Invest has predicted that by 2027, Tesla will be worth 11 times what it is today solely based on how valuable FSD could become! Such statements by Woods and Musk drive buying hype, pushing Tesla’s value way past where it should currently be. So Musk is right; without the self-driving hype, Tesla would be worth far less than it is today.
Now, Musk has also said he would be shocked if FSD isn’t already safer than human drivers and that FSD is already practically self-driving, with the driver only there for legal reasons. He has also said many times that FSD would only get more expensive over time as it becomes more capable and passes regulations to reduce the amount of human supervision it needs.
So, can you see why reducing FSD’s price now undermines Musk’s and Wood’s narratives and predictions about Tesla’s future? Either, Musk is exaggerating how good FSD is, or people are unwilling to spend thousands for self-driving cars. Now, there is data to support the first notion, as FSD customers only use the feature 15% of the time, heavily suggesting they don’t trust the software. However, the fact that Musk has to slash FSD’s price in half to increase income from FSD also indicates that even those who believe Musk’s exaggerated claims about FSD aren’t willing to buy it at Musk’s prices.
There are two arguments against this view. The first is that Tesla needs more FSD drivers to garner more data and develop the AI faster, and this price cut could help that. However, considering that over 400,000 Teslas have FSD in the US alone, I doubt that is the case. That is more than enough cars to give them the data they need.
The second is that this is the case right now, but FSD will get significantly better over the next few years, and this issue will be resolved. In fact, Musk has claimed that Tesla will be spending $10 billion on AI infrastructure in 2024 and that their upcoming Dojo supercomputer will be able to run millions of training scenarios for their self-driving AI, enabling FSD to reach better than human levels of driving automation. But there are three problems with this. Firstly, we can’t predict how larger training datasets will improve an AI, so there is no guarantee this push will equate to a significantly better self-driving AI. We are also getting to a point of diminishing returns, where the amount of training data required to reach a specific performance and predictability level spirals out of control. So, it is entirely possible it’s entirely unfeasible to create a full self-driving car with Tesla’s current approach. But the final nail in the coffin is that FSD only uses camera feeds to understand the world around it. This lack of redundancy in the system (i.e. no other sensors like lidar, radar or ultrasonics to fill in the blanks) means that FSD is inherently unsafe even if this AI expansion works.
But this fatal flaw in FSD didn’t matter. Tesla has been doing everything it can over the past few years to convince investors and market analysts that it has a super valuable product coming to market. This inflated Tesla shares, enabling them to lock in funding and grow their manufacturing at light speed, and even allowing Musk to secure enough loans to buy Twitter. However, FSD’s slashed price has caused a massive crack in Tesla and Musk’s smoke and mirrors. It has exposed the manipulation and possibly started a domino run to a market correction which will crash Tesla’s value.
Now, I’m far from 100% convinced this doomsday scenario is about to befall Tesla. There is a chance the AI training expansion enabled by Dojo will be significant enough to see enough improvement, and Tesla could adopt lidar and other sensors in its upcoming Model 2-replacing robotaxi to make it safer. But, from where I am sitting, this seems unlikely. As such, Tesla slashing the price of FSD could mark its downfall from an industry leader and stock market extraordinaire to an also-ran.
Thanks for reading! Content like this doesn’t happen without your support. So, if you want to see more like this, don’t forget to Subscribe and follow me on BlueSky or X and help get the word out by hitting the share button below.
Sources: The Verge, IBD, IBD, Observer, Inside EVs, The Verge, Electrek, IEEE Spectrum, Dan O’Dowd, Electrek, Tech Radar