Tesla will fail at Robotaxi! 

FSD 12.3.6 is taking me to work every day with no interventions. I'll trust my actual experience with this technology. It's easy to see robotaxis from here. Long TSLA.

robotaxi is inevitable. whether tesla is the first to solve it or not, they are coming.

Hilarious! My advice to you: put all your life savings in shorting tsla. You'll do great!

this man thought "end to end" meant get you from one end of town to the other, with no interventions... come on dude

6:40. Dude! LMAO!!!! There are 4 million Teslas on the road, that's 44 billion miles just for this year alone! Everyone of those Teslas have an FSD computer, running FSD in shadow mode, even if the software isn't enabled for personal use. EVERY SINGLE ONE is running FSD everywhere it goes. The only question is whether FSD is calling the shots or not, but it's always analyzing the trip for itself. 6:45. Yes, Tesla has data centers (not that Tesla "has" to, any more than OpenAI has to). Tesla is one of nVidia's biggest customers, Jensen Huang (nVidia's CEO) has talked about this. Where do you think these nVidia H100 AI training chips that Tesla is always buying from nVidia are going? To a junk yard? 7:24. You are so far out of your depth. They have FAR more than 1 billion miles. Also, they only look at 1 out 10,000 miles, because it takes 10,000 miles of driving for Tesla to encounter an edge case they haven't solved yet. That means 9,999 miles out of 10,000 aren't helpful because they have already been solved. This is in direct refutation of your previous points.

He has never driven with the old version. By his own admission.He has no idea how fast the new version has improved. The last few months have been remarkable! At this rate the only thing that will stop it are Govt. officials and Lawyers. It is better at driving than I am. They are no longer constrained by compute power. Keep up scooter. This has been the most difficult problem ever solved.

Uber and Lyft have to pay drivers. If you eliminate the driver, the cost is less and the market gets dominated by the company that has less costs and more profit. It's that simple.

probably the same dude said about apple will fail at iphone

Glad you shared your pearls of wisdom with the world. Please let us know how you feel when Tesla does not fail and you did not by stocks...

5:27. Funny when your opinion is wrong as soon as it was said. FSD is already having full trips without interventions all over the place, this is a documented fact. Do a little research. Just because it's not working without interventions in your specific area, does not mean it's "years away from working *anywhere*" when it's already working in many places right now. I've personally been in a Tesla where I was driven from my house to my mechanic, to pick up my Acura, with no interventions.

Im still listening but at the beginning you lost all credibility. TESLA used to be 50 bil Market cap end of story. They dont need to prop up their stock price.

Someone has a short.... lol...

My word, what a difficult watch. Another case of, "I'm an engineer, so I know better." Your claims at what Tesla is misunderstanding against you the enlightened expert are staggering. Your simplistic takedown over miles driven using crude numbers shows your complete absence of understanding of the technology involved, training, inference and control. Tesla doesn't need to keep billions of miles of video data. I am sure you're well intentioned, but I suspect you are blinded by your bias. I gave you a fair hearing and I did plan to list out and counter your many points. With the sheer volume of misunderstandings to address, I just can't justify committing any more time.

Just because there's questions doesn't mean there can't be answers

6:13. And you're an engineer who understands the core technology stack in the car, and you've performed a critical design review, and have come to identify unresolvable shortcoming in the fundamental design of the approach? No. You just see it doesn't accomplish what you think it should in your personal testing, and you have no greater insight than that. You can believe "the core is flawed," but that "opinion" is predicated on nothing than speculation based on very limit anecdotal experience. The real statistics gather over the fleet do not back up your assertion.

he said he never had it before but he can’t see the difference?

Just a few things: - The Robotaxi (RT) will cost Tesla less than 25k a piece to produce, once mass production is running, later on it will be less than 20k (not accounting for inflation etc.). - The RT battery will last for at least 500,000 miles, if not 1 million, the car will last for at least 2 million miles or 20 years. - The running costs per year for a RT will be less than 5k, while it will earn well over 10k, even if prices drop to 20 Cents/mile, at 50 Cents/mile it will make over 30k after expenses. - FSD is improving very quickly now, what took months before only takes about a week now.

You should Short Tesla…see how it pans out.

You seems having no idea about Tesla, or you plan to short Tsla?

Great points man. Thanks for giving us the perspective of a true driver.

This guys future is threatened by Robotaxi and is in denial

Yet in the real world, people make inteverntion free drives on RU-vid in various scenarios quite a bit.

You said you are an engineer. How come you are doing ride-share?

The guy earning $16 an hour on average appears to be saying he is smarter than Elon and knows best and what’s really going on. 😅

This just won't age well at all 😆

I think Tesla will sell shorts again soon.

Using your numbers. Av earnings $16.41 hr for rideshare. Robotaxi can work 20hrs a day (leaves 4hrs for recharge and cleaning between) = $328 a day. = $2,295 a 7 day week. You showed 5.5hrs a day (2000hrs a year) @ $40,000 for a robotaxi that's 17.5 WEEKS to pay for the car. Working life of 5 years = $500K profit split 60/40 with Tesla. $330,000 profit or $67k a year after the car pays for itself.

You have no clue what you are talking about in any way, shape or form...

Tesla will fix these issues and it will improve continually. I think. But the problems you are describing seem concerning and you make some good points and have some good questions. I'd like to hear what the answers are if anyone has them. I'd think FSD would have speedbumps and stop signs figured out bu now. Are others experiencing thses issues with FSD?

bro can you give us thoughts on neurolink next, too? lol go back to taking out the trash

Tesla doesn't have to analyze a billion miles. Just the edge cases and interventions.

Alright. You don't like it. So, don't buy it. It's still a free Country. Buy the other stuff and be happy with that. Simple.

Robotaxi is not good business in the first place. There is no first mover benefit. Only cheaper service wins at lowest profit.

Please go Short TSLA 😂😂😂

the value of a robotaxi with 50k miles on it that has already paid for itself is not $4k lolol it still has plenty of life left and can go earn you plenty more... where is your logic dude? these cars will depreciate differently. Honestly it just seems like you're afraid of your current situation being disrupted.

So we’re taking technical and financial advice from a full time Uber driver?? No offense to anyone. I respect anyone that works for a living, but if you have expertise in the financial realm you would probably Not be driving people to the airport to make a living…: just sayin’

"How....How...How....How....How....How....How.....How....Who...Who?" Oh you will see, you will see. Not all of them will be answeres right away with FSD. Some will be answered through the apps functionality and some through optimus. I would look for a different job if I were him.

Interesting vid. You get a few facts wrong but you obviously have many legitimate takes. FSD might be closer than we think, but the integration in various societies will take a long time.

Didn’t even bother watching the video… I will tell you exactly why Tesla robo taxi will fail. Long story short … it’s just not feasible for a long term strategy.. not how companies are ran today. Most of you are too out of touch of how much and the extent companies are willing to go through to save a dollar. You think all these semis on the road are using drum brakes in trailers cause they are the superior option? No, it’s so save money. This toxic save a dollar mentality extends to everything… the first that always gets the neglect is maintenance of equipment and fleet vehicles.. companies do the bare minimal. This will then lead to the inevitable which will be a failure on the vehicle cause they weren’t spending top dollar continuously on repairs and someone will lose their life. It’s only going to take a few lawsuits to send this idea to the grave. The reason so many of you are so content with this idea is that you see cars as a throwaway product. Companies will roll these cars into the grave before letting them go. The airlines WARNED US ALL about this tech… and even with their STRINGENT maintenance cycles we still have issues. Be real here people

You drove an incomplete product. That’s why there are no robotaxis yet.

Lame.

Hahaha. This will age well. Like milk

Not everyone is losing money driving rideshare, I think you need to look into that or get a new career

This guy knows nothing lol

No datacenter? Why are Tesla NVIDIA's biggest client?

if Musk is bad, what is ford, gm etc. complete false transcript

Clown show. This guy has no clue!

Also… Mr Expert: there is no need for a “special car” to be built. Any current Tesla would be fully FSD capable once the software is fully functional. Wow this guy is clueless. I feel dumber as a result of watching this video.

The comments you have are fair but to say they will fail is a complete misunderstanding of the technology. Do you understand machine learning and neural networks? It does not sound like it. It’s not running a “program”. It’s cycles through decisions. Improvements requires more training. The results are inevitable with more training.

Brave enough to dislike this type of BS.

Yeah, you are screwed. No more work as a driver, better prepare while you still have time.

You Fail

You have no idea what you're talking about at all. And that's sad coming from a Tesla owner.

You bring up a lot of interesting points. Heres where I agree and disagree. Agree - The economics of the Robotaxi are not going to be as promising for customer vehicles as people think. If its going to be profitable at all, it'll mostly be for Tesla once they ramp their own fleet vehicle. Competing against Uber and Lyft is going to be difficult but without how cheaply the next gen platform of the robotaxi will be, I woudn't be surprised if they dropped the cost really low because they will have a much lower cost of building the vehicle so the amount that they'd need to profit from each vehicle is a lower bar to clear. The real problem here is that Tesla will essentially be competing with their own customer base for rides so the idea that this will be a stable income for Tesla owners is far fetched to me. - The ergonomics of keeping the cars clean, charging them, and handling edge cases where humans are needed is going to be a problem. The only thing is that this is a problem for any company that uses cars with no driver. There may be software fixes for these. Internal cameras might detect dirtiness in a car which will then direct the car to a service center while a clean car will pick up the passenger. The same might go for charging where they autonomously go to charging stations that are manned by Tesla employees once they are logged out of the fleet until they are deemed suitable for tasks. Tesla might force gate operators to use some open standard that allows their car to communicate with the gates to open them. - The human aspect like helping the elderly, hailing rides to the right location, etc will likely keep humans in the loop for the foreseeable future. People tend to think that competition with new tech = old tech completely wiped out and I don't see it as such a zero sum game. Sure autonomy might progress to the point that its safer than human drivers but I still think that many people will find it preferable to ride with humans for a long time. Cost to the consumer is going to be the hard part. But I don't think any of this is going to bode well for how rideshare drivers are paid. Lyft and Uber will likely respond by cutting your pay before they cut their profits. So while I do think that human drivers are going to be desired, the corporations might opt to make it less economically viable for you to use their services as a stable source of income going forared. Disagree - The points you made about data collection are misguided at best. Tesla is logging your drives. If you have 1m cars on the road and they each drive 10 miles a day, thats 10m miles a day that could be collected which translates to over 3.3B miles a year. Now I don't believe Tesla logs all drives from start to finish but its not hard to understand how they have billions of miles stored, especially because they've been recording data for a decade and they have a much bigger fleet than 1m cars on the road. Even if they didn't have their own data center, they could be storing the data on cloud storage. There are plenty of ways to store the data. This is not the same as a data center that is focused on training on that data.You can transfer over the data that you want to train on. Where the data is stored on and where the data is trained on don't need to be the same compute node. - Stating that FSD doesn't work is missing so many things. You might notice a few bugs in the current version but this negates the fact that they've been constantly improving to get up to this point. I remember seeing FSD struggle to get through roundabouts. Now it does them perfectly. It also never used to be able to do a u-turn. Now it does them regularly. Neural networks are function approximators, meaning they will never give 100% predictable outputs. It'll take time and a lot more training to iron out bugs in the software. But this is an issue of time and training compute, not capability. FSD will continue to have problems, but the point is that neural networks will eventually be able to solve them more and more over time until its capability can exceed human driving. Neural networks have superseded human ability in many domains with enough training. Driving is much harder because the amount of things that it needs to account for are near infinite, so it takes much longer to achieve but it will eventually surpass human capability. No one knows when that is and its easy to have overconfidence in them which is why Elon has had so many bad predictions on when this would happen. So while its easy to write what Elon said as a lie because it didn't come to fruition, the reality is that neural networks take a lot of time and effort to get to function the exact way you want them to when it comes to complex problems. They're not as simple as tuning once and then the problem is solved. For a task like FSD it will take many petabytes of data and a robust architecture in order to solve properly. You mention OpenPilot, they also are using neural networks in the same way to solve this. To really measure who's more ahead, you have to see how well the software performs in different environments and we don't have that data for OpenPilot as much as we do with Tesla because of how many people record their rides for us to evaluate.

🥳 ... LOL...

Easy to understand why you only have 9 followers

Haters gonna hate😂

Elon never fails!

He has done some impossible shit. Short seller 🤫