The Tesla Self-Driving Myth Dies Another Death – Tesla Motors (NASDAQ:TSLA)

The 2018 Cadillac CT-6, which includes the Super Cruise option.

I’ve already shown how Tesla (NASDAQ:TSLA) massively lags other companies when it comes to self-driving technology:

All of the above were and are very evident. Yet, believers kept on believing in something else: that Tesla was ahead. How could such obviously false belief even hold? The reason is simple:

  • The overwhelming majority of people do not follow and are not familiar with the actual state of the art when it comes to self-driving.

Instead, most people, and certainly the Tesla faithful, were familiar with something else. They were familiar with Tesla’s Autopilot capabilities. Those driver assistance capabilities, most notably the lane keeping abilities, were indeed the best which were commercially available.

As a result of this, most people have wrongly thought that Tesla led in self-driving technology. They thought that driver assistance leadership was the same as leading in self-driving technology development.

It was not, though. The best of today’s commercial driver assistance efforts is orders of magnitude less capable than any of the leading self-driving efforts being developed by Google (NASDAQ:GOOG) (NASDAQ:GOOGL), General Motors (NYSE:GM) or Nissan (OTCPK:NSANY). Those cars navigate city streets autonomously, and they do so safely under trying conditions. Google’s cars can go for thousands of miles between needing human intervention, and GM’s can go for hundreds of miles even on difficult San Francisco streets. They were able to do so on average during 2016. They’re likely a lot more advanced now.

So what does Tesla’s new self-driving death consist of? It consists on the realization that Tesla is about to lose even its driver assistance lead.

Whereas before Teslarians (Tesla fans) could point to Autopilot and say there was nothing quite like it, that’s really about to change – immediately. For instance, just recently the trusted Car and Driver magazine reviewed several driver assistance systems. Among them was Tesla’s Autopilot 2.0 and Cadillac’s Super Cruise.

It’s evident from the review that Cadillac’s system is the superior one for the purpose. A few of phrases make that evident:

Park it in the right lane, though, and Super Cruise could seemingly track forever. In 40 miles of driving, the system didn’t ask us to take the wheel, except when we intentionally aggravated it by looking away from the road for an extended period. After less than 10 minutes, Super Cruise’s rock-steady competence became boring.

This comparison will likely be made over and over. Some won’t quite understand immediately why GM’s system cannot but be superior, for instance, Jalopnik thought the two systems were more or less equivalent. Jalopnik even wrote about how Tesla has automatic lane changes where the car checks for safety while GM does not.

Alas, as was Charles Mackay once said:

Men, it has been well said, think in herds; it will be seen that they go mad in herds, while they only recover their senses slowly, and one by one.

Hence, above, Car and Driver has recovered its senses, and Jalopnik is in the process – just by deeming the systems more or less equivalent. I’m saying this because:

  • Tesla’s system remains highly unsafe since the car is prone to have a mind of its own and simply track “the wrong lanes.” Meaning, the car can suddenly swerve out of its lane or into a divider. GM’s car won’t do this (provided it’s not used in a construction zone, where the Tesla can also fail). Why not? Because GM’s solution doesn’t just rely on seeing the lanes. The car uses an HD Map of its surroundings and has precise information about the road and lanes it’s traveling on, as well as its own position even without being able to see the lanes. As a result, the car does not need to guess. Also as a consequence, it won’t swerve suddenly because it thinks the road/lane follows this or that curvature … since it knows what the curvature is and where the lanes are, at all times.
  • Tesla’s automatic lane change previously couldn’t check for safety (the range of the ultrasound sensors was not enough), and now likely doesn’t check for safety other than using the improved ultrasound sensors (whose range is still insufficient). It likely doesn’t because Tesla hasn’t enabled the use of the required cameras yet – though this might happen in the future. As a result, using automatic lane changes without the driver checking for safety is still unsafe in the Tesla. The same care, obviously, remains needed in the Cadillac.

Anyway, the relevant point here is that Cadillac’s system will be shown to be superior for the purpose – providing reliable driving assistance in highway driving. This will slowly filter out through multiple reviews. Tesla’s system, though, remains unreliable and prone to suddenly change course.

Both systems require the driver to keep his attention on the road, of course. But Tesla requires the driver to be on the alert not just for what the road might bring, but also how the car might irrationally behave when nothing threatens it. GM’s system will just require attention to what the road might bring.

This difference is massive, and those reviewers knowledgeable enough will identify it: What the road might bring is predictable if you are looking at the road. You know when you can be relaxed since there is no immediate incoming danger. Tesla’s approach, on the other hand, means action might be required at any moment, out of the blue, even with no observable danger within sight, just because of the car’s reactions to things the driver does not perceive as problems. This is intrinsic to the technological approaches followed by GM and Tesla.

Conclusion

The myth that Tesla was ahead in terms of self-driving technology was long dead for any knowledgeable person following developments in the industry.

However, for the general public, the myth was and is as alive as ever. This happens because the general public doesn’t follow the state of the art when it comes to self-driving. Instead, the general public conflates the state of driving assistance with self-driving. And Tesla did indeed lead in commercially available driving assistance (the lane keeping part, at least). As a result, the general public thought Tesla was the leader in self-driving technology.

It is that second myth that starts dying now, with Cadillac’s fielding of a superior driver assistance feature due to technological features not used by Tesla. The myth also will likely take further blows as other advanced systems make their way to the market.

Of course, the myth certainly won’t survive past the day where Google and GM start providing services based on self-driving technology even while Tesla has nothing to deliver.

In the meantime, rumor has it that at least 35,000 Tesla owners bought the FSD (Full Self-Driving) option for their cars. Tesla won’t deliver on this ability (not before others, and not with the current hardware, anyway), and these 35,000 cars (and growing) will become a giant liability.

In my opinion, Tesla will try to fend off the liability by returning the $3,000 option value per car to the owners, but this won’t be sufficient. The reason is simple: those owners can easily argue that they just bought the whole car because in time it would self-drive itself. As a result, the liability will be for the whole car and not just the FSD option.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a short position in TSLA over the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Leave a Reply