Disclaimer: I am bullish on Tesla and have been since I started writing & investing in Tesla in 2013. And Elon reads my essays so there’s that. I would also like to think I am fair in my critique and that I am not just completely blinded to think Tesla is infallible and can do no wrong.
Both sides of the Tesla community, the bulls and the bears, can be quite passionate and often times wrong about what they believe and spout off to others.
I am quite critical of the #TSLAQ bears as they seem to think that Elon is a charlatan and that Tesla as a company is worthless, fraudulent, etc. Clearly these people are not being reasonable as an objective person can see that a million free human beings (and counting) choose to purchase Tesla vehicles through voluntary exchange. Objective people can also see that Tesla has cutting edge microchip, AI technology & factory technology, all of which is tangible and clearly not fraudulent. Finally, Tesla is continually audited by PricewaterhouseCoopers LLP, which means that Tesla bears by extension believe that PricewaterhouseCoopers is in on the fraud. [edit: a very smart woman corrected me and pointed out that Tesla could technically be duping PWC by providing false data. How in the world is 3rd party FULL auditing back to the bank account not yet the norm? I bet the blockchain will change that!] Tesla bears simply refuse to acknowledge these facts (willful ignorance) and never look at the past to see exactly how silly they sound to other adults given the benefit of hindsight and historical data.
That said, I feel like a lot of people in the Tesla bull community are also blinded to a few simple realities, mainly that no company, including Tesla, will provide Level 5 Autonomy (aka Full Self Driving) any time soon. Not this year. Not next year. Not in 3 years. At this point, owning and driving a Tesla with autosteer myself, I cannot see Level 5 Autonomy being achieved any time in the next 5 years. And it’s likely not going to happen within 10 years.
When a loved ones daughter was born eight years ago I predicted she would not need to get a drivers license. I am ready to walk back that prediction and say that I was wrong. The ability to use a vehicle with no steering wheel in it, complete level five autonomy, is nowhere close. I own a Tesla. I use autosteer daily and I admit the technology is absolutely incredible. But any objective human being should be able to see that there are just far too many edge cases to train a computer to understand to be comfortable putting your loved ones in a vehicle that does not have a steering wheel.
On a recent drive in my Tesla Model Y there were multiple random pieces of tire in the road that required maneuvering around, and shortly after something came flying off the back of a truck right at the car. I made a split second calculation that it was a harmless piece of paper and did not swerve my vehicle. However if it was a piece of metal or piece of plastic or something I might have had to swerve. It is going to be a very long time until the neural network is sufficient in training artificial intelligence to mimic those reactions 100% as good as a human let alone 1000% as good as a human.
I believe that it will be a long time before computers are even equal to humans in all cases of driving, let alone magnitudes better. But unfortunately I see a lot of the Tesla bull community dig in and act like a Tesla artificial intelligence is more advanced than it is. These bulls are either naïve, willfully ignorant or simply repeating things they hear without doing any critical thinking. I would guess a decent number of them do not even own a Tesla vehicle to be able to put their money where their mouth is.
Tesla’s capabilities with regard to full self driving is quite a polarizing topic. Just the fact that they call it full self driving when it is obviously very far from full self driving is enough to get the Tesla bears going. Bears will correctly point out the ridiculous misnomer however then they believe at all of their other critiques of Tesla are valid when most of the critiques are anything from ill-informed to flat out wrong. The vast majority of Tesla bears have often never ridden in or driven a Tesla and they simply dislike Elon Musk or like being part of a negative community or both.
(Sidenote: If you want to explore an interesting phenomenon go down the #TSLAQ rabbit hole someday. You will find an entire community of people who seem to base their value as a human being around their perceived fraud of Tesla and its CEO. Some #TSLAQ allow the hatred and negativity to consume every waking hour of their day and unfortunately these people probably do not realize how much it is a waste of their life. They would be much better off replacing that negative energy with something that they love and enjoy and never thinking about Tesla ever again. It blows my mind that these bears will literally tell me that the car I willingly chose to purchase through voluntary exchange is bad for me even though I own it and drive it and love it.)
Tesla screwed up by choosing the name “Full Self Driving” for a baby version of the product and then they seem to have dug in when they must realize that level five autonomy is a long long time away, even if Tesla does not say so publicly. Because Tesla fails to address this poor naming, bulls and newscasters do things like ask questions such as “will level five autonomy be here by the end of the year (2021)”. They are actually serious with this question and ask it because Tesla and its supporters seem to give off the impression that level 5 is not far away. THIS IS NUTS! LEVEL 5 AUTONOMY COULD BE 5-10 YEARS AWAY, OR MORE!
It is true that Tesla has made tremendous progress in autonomy and every day advances it even further but it is also true that they are a long ways away from taking the steering wheels out of our cars and they would be better off just being honest and telling people that. Honesty would quiet down the bears and help make the bulls look less stupid when bulls act like level five autonomy will be here “any day”. Autonomy and Tesla making an INCREDIBLE electric vehicle are not mutually exclusive. The incredible vehicle can (and does) exist RIGHT NOW, regardless of any autonomy. The beta version autonomy that we have at the present is just a bonus and will only get better from here. But it take some time to reach Level 5. So sit back, relax and enjoy the ride (with your hands on the wheel…for now!)
Discover more from logical thinking in an illogical world
Subscribe to get the latest posts sent to your email.
Interesting article. I didn’t catch if you personally have FSD Beta currently. I think that’s probably important for one to make a well-informed assessment of it. I own it and have been using it for around 1.5 years or so. In fact, I just got home after a ~7 hour round-trip drive with FSD Beta doing 99% of the drive very well. Based on my 1.5 years experience with the product and seeing the progress each month or so and a couple decades as a software dev and tech enthusiast, I predict Tesla will be ready to start testing FSD for robotaxi service in some non-trivial capacity in 2-3 years. I think it may **appear** to be Level 4 to most casual observers by then end of 2023.
I believe this because it’s nearly capable of robotaxi scenarios right now to a surprising degree. Some of its biggest remaining issues I can think of off-hand are:
1 Doesn’t consistently select correct lane to be in in a timely manner
2 Doesn’t place car in most ideal part of narrow lanes
3 Still needs to estimate speed/acceleration of other vehicles with more accuracy
4 Needs to assess when it’s safe to go (mostly at intersections) more quickly
Based on rate of improvements I’ve personally experienced, I’ll be surprised if #1 & #2 aren’t fully solved within 12 months. 1 has improved dramatically just in the last two months with very specific and intentional re-architecting of the machine learning networks. For #3 & #4, in the absolute worst case scenario, they can throw more compute at the problem in the form of the upcoming Hardware 4 computer which I’m 100% certain would do it. But I’m certain the existing computer will be able to do it too. The irony of software is that early versions of software are always the least efficient and therefore require the most compute power. Subsequent versions (with same features) become much more efficient over time and need less compute. So, speeding up #3 & #4 can be a function of compute and/or a function of code/ML network architecture/data optimization.
Regarding the naming – the oddity for the market is that we’re not accustomed to seeing products that aspire to be a thing, released to the public before they’re done becoming that thing. I think that’s what’s confusing some people. It is appropriate, in my opinion, to call a product that is absolutely being designed to enable full autonomy, Full Self-Driving. I also think it is appropriate to slap the name “Beta” on the end to convey that it is incomplete. However, I’m a software professional and fully understand the meanings of these terms and the thinking. I get that the general public may not fully get that or appreciate it. I think ultimately, this will be a moot point.
LikeLiked by 1 person
Thanks, Terry! I am in the FSD Beta program on 2 LRMY (2020 and 2021) and I agree with #1-4 as you state. Just not obvious to me that the timelines will be as short as you and others predict. I certainly could be wrong, I often am, and I hope to be in this case!
You are correct on the name, but combined with most people not understanding the “beta” part of it, and Elon’s ridiculous timelines (NYC to LA in 2018! or whatever it was) the naming could have been better or at least dealt with by using a little more of a subtle touch.
I do hope that as testers we get HW 4 for free/reduced rate. I really think the camera upgrade will make a big difference. In addition, I hope Tesla stops being so silly and adds back FLIR or similar (see my other writings).
LikeLike