Photo: Tesla Motors
Starting last week, every new Tesla car that comes off the assembly line in Fremont is going to be equipped with the hardware necessary for Level 5 autonomy. Those are Tesla's words, not ours.
This means that, in theory, we probably already have a few privately-owned cars on US roads that are capable of driving themselves from coast to coast, as Elon Musk put it. Take a moment to allow it to sink in, as this is a significant milestone in the history of the automotive industry.
As big as the news sounds, Tesla's stock price has actually taken a dip after the announcement, dropping by 2 percent on a day that was supposed to lift the company above any other competitor, at least regarding its advancement in self-driving technology. But that didn't prove to be the case, and one might wonder 'why?'
The answer to that question has to do with Tesla's approach toward the safety of its clients. After initially admitting that the Autopilot was essentially in beta testing, Tesla changed its rhetoric after the Joshua Brown incident in May which showed what can happen when its users take things too far, but it also highlighted the huge limitations in the system.
Elon Musk, however, is adamant that the Autopilot should be encouraged, to back up his obviously biased claims, he uses some statistics that prove mathematically how Tesla's system is better than human drivers in regards to the number of deaths each of the two cause over a number of miles.
Don't get me wrong, I'm not that naive: I know he runs a business and needs to make a profit, but that doesn't mean he couldn't have the Model 3 sell only with the feature installed. If the Autopilot is indeed so good, then that move alone would save plenty of lives. But it would also do something else: with the suite retailing for $5,000 or $8,000, it would raise the Model 3's starting price over the promised $35,000 mark, and that's not good for marketing.
Well, be that as it may, the Model 3 - just like any other Tesla built from now on - will come with all the necessary hardware installed. That means cameras, radar, and sensors, all working at full capacity, but not for the owner of the vehicle, but for Tesla. They will run in 'shadow mode' collecting data and comparing the reactions of the human driver to those that the system would have taken in the hope that over the course of one year, they will have gathered enough data to prove the AI would be a better driver than the current owners of the vehicles.
I have no doubt that Tesla will eventually do it. Maybe not as fast as Musk would hope, but it will get there soon, and it will be before any other major carmaker. Will there be any fatal accidents in the process? I'm pretty sure everybody hopes there won't, but not everybody thinks that as well.
As big as the news sounds, Tesla's stock price has actually taken a dip after the announcement, dropping by 2 percent on a day that was supposed to lift the company above any other competitor, at least regarding its advancement in self-driving technology. But that didn't prove to be the case, and one might wonder 'why?'
The answer to that question has to do with Tesla's approach toward the safety of its clients. After initially admitting that the Autopilot was essentially in beta testing, Tesla changed its rhetoric after the Joshua Brown incident in May which showed what can happen when its users take things too far, but it also highlighted the huge limitations in the system.
Elon Musk, however, is adamant that the Autopilot should be encouraged, to back up his obviously biased claims, he uses some statistics that prove mathematically how Tesla's system is better than human drivers in regards to the number of deaths each of the two cause over a number of miles.
Using cold numbers
The thing is, using cold numbers when talking about lives isn't exactly what the public is used to, and there's another aspect that doesn't help with Mr. Musk's credibility: if he's so worried about saving lives, why is he charging extra for the Autopilot?Don't get me wrong, I'm not that naive: I know he runs a business and needs to make a profit, but that doesn't mean he couldn't have the Model 3 sell only with the feature installed. If the Autopilot is indeed so good, then that move alone would save plenty of lives. But it would also do something else: with the suite retailing for $5,000 or $8,000, it would raise the Model 3's starting price over the promised $35,000 mark, and that's not good for marketing.
Well, be that as it may, the Model 3 - just like any other Tesla built from now on - will come with all the necessary hardware installed. That means cameras, radar, and sensors, all working at full capacity, but not for the owner of the vehicle, but for Tesla. They will run in 'shadow mode' collecting data and comparing the reactions of the human driver to those that the system would have taken in the hope that over the course of one year, they will have gathered enough data to prove the AI would be a better driver than the current owners of the vehicles.
I have no doubt that Tesla will eventually do it. Maybe not as fast as Musk would hope, but it will get there soon, and it will be before any other major carmaker. Will there be any fatal accidents in the process? I'm pretty sure everybody hopes there won't, but not everybody thinks that as well.
Google News