Tesla's Robotaxi Network Launch
HTML-код
- Опубликовано: 14 май 2024
- Larry Goldberg is a serial entrepreneur (co-founder of Sapiens Decision), angel investor, and co-author of the book “The Decision Model”.
Follow Larry on X: @TeslaLarry
My website: www.TeslaInvestor.com
Get Free TESLA Milestone Tables
Check out 15+ modules of resources for the $TSLA Investor
Join this channel or Patreon to get access to perks:
Get free access to 15+ modules of TSLA investor resources
Become a Member: / @brighterwithherbert
Become a Patreon: / brighterwithherbert
Use my referral link to buy a Tesla product:
ts.la/herbert23392
❤️ Thank you to everyone who supports me and this channel! I really do appreciate your likes, subscribes and comments.
Let's get brighter!
Website: www.herbertong.com
X: @herbertong
Disclaimer: All content on this channel should not be construed as professional financial advice or recommendation to buy or sell any securities.
Author is invested in multiple Humanoid companies and is long TSLA and other stocks at time of original video publish date.
#Tesla #TSLA #stock #ElonMusk #investing #teslabot #technology #ai #TeslaNews #TeslaFSD #teslastockanalysis #TeslaUpdate #ElonMuskNews #EV #FSD 
Наука
I hope this video helped you get brighter about Tesla!
My website: www.TeslaInvestor.com
See the MOST comprehensive resource for the $TSLA Investor
Get FREE TESLA Milestone Tables
Join this channel or my Patreon to get access to perks and free access to 15+ modules of TSLA investor resources
Join this channel: ruclips.net/channel/UC4DBLlq1x0AKmip1QJUcbXgjoin
Join my Patreon: www.patreon.com/brighterwithherbert
Use my referral link to buy a Tesla product: ts.la/herbert23392
Thank you for supporting this channel! I really do appreciate your likes, subscribes and comments.
Let's get brighter!
Herbert
Thank you Larry. May you live long in good health.
Long … so that you see the outcome of all your judicious predictions and insight
I agree with Larry!
The introduction of a non-robotaxi Model 2 prior to robotaxi would help tremendously with training and acceptance by regulators.
Herbert better asking questions than answering
He shouldn’t interrupt Larry when he’s answering….its disrespectful
I agree with Larry 100%. Robo-taxi will be a very slow ramp. Some big markets like Europe wont allow them any time soon. I dont see why tesla cant do both the consumer driven version and taxi at the same time.
Energy energy energy!
no growth in 5 Qs ...
A possible event where a lot of the Tesla community set each other up, and drove FSD in a huge fleet at the same time in a smaller town / limited area in a big city / at night. Then the world could open its eyes to how fantastic FSD is close to becoming 😍 Aren't you the man to create such a party 😉
It’s been “close” for years.
We DO need the 25k vehicle to come in the next 2 years if we want the stock to go up...thats the reality
No its not. If it was the correct move Elon would have done that. He has a track record, all you have is broken record.
@@michaelnurse9089 hope u right man cuz am heavily invested !
Sorry. If elon was always right, the stock price would not be where it is. There wouldn't be an issue with delaware; and v 12, semi and cybertruck would have put us in record territory. e@michaelnurse9089
@@michaelnurse9089a little extreme, no?
Elon is a more accomplished human than I could hope to be but that doesn’t mean he is flawless. The fact remains that until $25K model ramps and FSD is ready the 3 & Y are going to suffer price cuts, market saturation and also have little room left to save on costs to build.
This would be less of an issue if $TSLA didn’t have a PE ratio that implies earnings growth this year even at $150 (and yes I do believe in FSD, but we’re trading with wall st etc who will not price it in the same until it is a present reality)
The 25K vehicle is sorely needed, especially with the competition emerging in China.
Long term Tesla bull here.
The cyber truck was a mistake . Should have produced the model 2 to introduce an affordable electric car to American public. Of course you would have to have good range and cold weather charging capabilities which would require different battery chemistry. Instead of the cyber truck which doesn’t look like it’s going to appeal to most truck buyers
Larry rues on all things Tesla - I say that as a Tesla investor!
Cheers guys
Hopefully Robotaxi comes to Minnesota twin cities.
Uber and Lyft are pulling out of the area, as they don't want to pay drivers higher wages.
People don't realize how good FSD is now. What people don't realize even more is how far & rapidly FSD will progress over the next 24 months. Robotaxi / M2 will come from the unboxed manufacturing in the next 2 year period coincident with improved FSD. Tesla is anticipating the advances in AI / compute / improved manufacturing and matching Tesla's business future to the tech progress. Energy business will be profitable during this time. Tesla needs to go into nuclear to ensure power availability for the future AI & data centers.
Tesla does that now with the Super Charger Network. They know how many cars are heading or re-routing to chargers and open stalls. So Robotaxi cars will know all that now and I can see it on my Screen in my Tesla car while conditioning the battery. Tesla has their own “Air Traffic Control” live and functional. Great job guys!!
good news Wallstreet will drive stock down to buy!
Herbert, Buffalo is in New York state. Lol!
Wireless charging
Plan B: ramping up production of a Model 2 with FSD Hardware to which Tesla can sell FSD Software or a subscription for private use as they should do for every other capable Tesla car out there.
There is one thing people REALLY need to understand. Very little mega successes were highly anticipated. What I'm saying is is that when you announce your target and roadmap like this. Very few companies can keep the competition off their back. As you as a company get closer, so does the competition.
Data is the moat. If they start now they will be 7 years behind. It will take them 3 years to start so minimum they are 10 years behind. They must also commit huge capital and the other car makers outside China have no access to capital, with share prices in the toilet.
It's not so simple. Just like building good BEVs wasn't as easy as everyone thought.
RoboTaxi might be too difficult for any real competition.
But I expect BYD to license FSD and they could then do RoboTaxi in China, and wherever they are allowed. So not the US, of course, but everywhere there's a free market.
I think the Chinese are the only ones who can compete!
Could the robo taxi be a 25k 2 seater for sale to individuals as personal vehicles?
Herbert, "wall" is singular
Thank you I thought it was just me 😂
I think the 25k car first, then the robotaxi when FSD is really ready. It is too risky now and there will be too many setbacks. The company mission is to accelerate the transition to electric cars. A very appealing 25k car that people will want to buy above all others is what's needed. Keep the factory going and employees happy. Also, everyone assumes riders will be cool with a driverless car. I think many would be terrified. And with no driver there will be plenty of mischief in the back of every variety. Climbing in the backseat may be an adventure in biology. 🤣
The $25k won't be unveiled until it can be produced at scale!
That's why they focus on the RoboTaxi now. It's one platform, so if you ramp it up with one vehicle, you can do the others pretty easily.
When the platform production is working well, they will reproduce it to scale up. Then they can build the $25k car at scale.
Otherwise, they would just create unreasonable wait times!
The moment The owner of a full self-driving capable Tesla finds out that they can generate crazy money there's nothing stopping hundreds of thousands of cars joining the robot taxi network the moment it's available and working.
IDK what you define as crazy money but operating a taxi is fairly low income
Tesla will take most of that money. Owners will get 20-30%.
I don't it, there so many rule and regulations they need follow.
@@deltaechomusicnh555 Agreed. Insurance, and making sure the Tesla is mechanically sound. And there will be those days where your Tesla comes back with vomit in the backseat.
@@hankkingsley9183 That's due to the high cost of paying drivers. $100k at least per year to keep drivers behind the wheel for 18 hours per day, 365 days per year.
Many people are attatched to thier car and will never share that.
You don't need to wait until the next-gen cars are built to start the robotaxi network. You just need FSD working fully autonomously in the current generation of vehicles. So, all of the regulatory approvals, insurance, payment system (just need a tesla account and the customers will be billed in exactly the same way as current Tesla owners are billed when using Superchargers), phone app (Tesla already has one of these) chargers, promotion to the public, and other infrastructure and marketing can be done before the actual robotaxi vehicles roll off the production line. So, all of the support infrastructure can be in place before the dedicated robotaxis are built and the business can get up and running quickly by using the current generation of vehicles.
100%
Thank god somebody else gets it! FSD is the key, having a robotaxi and no software is a waste of time. If you have the software you don't need a dedicated robotaxi any M3 or Y will do. You don't need to have drivers in the taxi and have some trial period because you can show regulators your FSD data. I would pick Austin and roll out 1000 cars as soon as the FSD software is stable. Oh, if you are just now thinking you need to make an app....Fire yourself, that should have been 90% years ago.
And don't need no UBER data.... Waste of time and money.
Tesla isn’t getting level 5 autonomy any time soon and even when/if they do they will then have to clear some major regulatory and legal hurdles. In the meantime and given all this will need massive investment where does the money come from?
@@tatata1543 Massive investment in what? Insurance? A phone app that they can probably roll out in a couple of weeks? Chargers? They will start the robotaxi service in locales where they already have sufficient chargers to support the service. And they currently have about $30 billion dollars in cash. That employs a lot of cleaners who can vacuum the cars while they are charging.
Larry is correct. People that need transportation today don’t by cars based on the resale value. Herbert is correct if he replaces Robotaxis with FSD or autonomous vehicles. During the launch of the Robotaxi service, sales of Tesla’s current models of EVs should increase because of the availability to drive themselves. A lot of people that are not interested in enlisting their EVs on the Robotaxi network, will want to buy an EV with a built-in chauffeur or chauffeur-capable. Many people overlook this value to many customers because of their hyper focus on the Robotaxi service. Some people buy Teslas because they don’t want to drive or like to drive. Others would want the ability to subscribe to FSD a few times a year for long trips or when visiting area they aren’t familiar with. These people might buy a Tesla just because of this capability alone.
Will Farzad’s idea that Tesla will sell the shell and the current fleet operators will customize it accordingly … like Airbus or Boeing that sell the airplane and each airline personalizes it for the usage made …. Fun, business, events, …. 1000x possibilities!!!
There will definitely be the need then for robot arms that automatically fill up the Robotaxi batteries … similar to the snake arms that plug the charger into the car charging connector ⭐️👍💥🧨⭐️🔥
Lastly need a service provider for private owners that will cover prepping of the car for a fee!!👍👍🚀🚀
Tesla does not have to make a loss on Robotaxi first year.
Fleet operators will emerge and snap up all available Robotaxis. Tesla can sell them with a massive markup including the FSD license.
Later as they ramp up, Tesla can keep a higher and higher percentage of vehicles to add to the network themselves.
I also agree with Herbert that demand for all Tesla vehicles will go up when FSD becomes level 3 or 4. Resell value will go up. It’s not about individuals. The market will push the price up.
Once again Larry is absolutely correct! If Tesla teams up with Uber and Lyft as companies or with the Uber and Lyft drivers, they can gather the ride hailing data to FSD for Robotaxi service. Uber, Lyft and taxi drivers make dozens of decisions daily that a typical Tesla owner never encounter. If Tesla provides Uber, Lyft and perhaps taxi drivers with a free or highly discounted lease of a Tesla today, these drivers will gather the ride hailing driver decision data to train FSD. At the end of this “Beta” Robotaxi launch, Tesla could offer these soon to be disrupted Uber and Lyft drivers, steep discounts on the sale or continued lease of these vehicles, thereby reducing the blow of loosing their jobs by providing them a means of income without their Tesla vehicles on the Robotaxi network. This act by Tesla would help to reduce the opposition from the Uber, Lyft, and taxi driver’s lobbyists.
Will not including FSD with compact car help avoid Osborne effect with 3 & Y?
Absolutely agree with Larry. We need the 25k for sale at the same time. Magic puffy thinking is not going to make the robotaxi any better.
Uber won't oppose RoboTaxi!
They will just convert itself into a RoboTaxi company.
Just like Uber today, but without having to pay drivers.
Of course they are planning for this!
I guess they will start using Teslas with FSD supervised in some locations to confirm that it's working.
From this, they will be well prepared to get regulatory approval for going driverless.
They might start with remote communication with the passengers, because Grok won't be ready for that.
Tesla should discuss this with Uber, it's a win-win!
Agree. Uber would be the perfect place to get the data needed under supervision to prove to regulators that fully autonomous FSD should be permitted. Although the data from Tesla customers could be sufficient. And while Tesla probably will operate its own taxi service (and allow Tesla customers to participate) running Uber out of business would be a very bad move that would likely trigger govt anti-trust regulators.
There could be substantial pretty-soon money when current Tesla owners rent FSD. Jeannine
My FSD wouldn't drive in the fog yesterday. It kept disconnecting and telling me to take over. The visibility was such that was totally viable to drive if you slowed down a little and payed close attention. Tesla needs to fix this. Robotaxi can't shut down in medium fog.
I think Larry is a little skeptical about the Robotaxi and so am I.
Larry isn't scept8cal about Rob9Taxi. He is just tempering the expectations about the speed of the rollout.
That's just being realistic!
The first year won't even show much profit, because of the ramp up costs, and the cost of building up the taxi service.
But it will show that RoboTaxi works, and that should be enough to convince the few smarter investors.
A repeat comment. I bought a 2012 model S. I'm a techi. The next time you are at the grocery look around. How many of those people will ride in a car without a driver?
James Douma is a true polymath.
Uber reported its first profitable year in 2023, after 15 years of operations.
It takes time.
Uber doesn't compare to Tesla
If you don’t have to pay the drivers it is easier to be profitable
Thinking about Uber and Hertz. What value can they really add to the RT business? RT customers will expect a clean, convenient, reliable, safe, and timely service. Maps? No. I'm a part time Uber driver. I use both Tesla/Google and Uber maps. Both have good and bad aspects. A wash. "Uber knows where the customers are." Yes, but so does any driver past his first month driving. Dispatching software? No, Tesla is a great software company and besides, the two apps are different. Uber is a service that arbitrates scare drivers. RT is a fixed price business with two price bands, peak and non peak, on a published price schedule. The dispatching algorithym - pick up the closest customer call. Hertz is likewise a non-value add. Tesla may not want to do it, but quality will demand that Tesla own and service its own fleet.
Thanks for the discussion.... Thinking through the Robo Taxi "back-office" structure. Simply, outsourcing could bring un-intended consequences. Those who only get paid per car,, will find a way to do less to increase the number processed. Larry, I feel, is correct. Don't get enamored with the FSD technology and think it is DONE. It's the structure that services customers. Thanks again.
Speculation. M2 will be the same car as the robotaxi. Except for M2you will be able to remove the steering wheel and pedals because it’s all control by wire. Add an airbag into dash once steering wheel airbag is removed OR it comes with dual airbags. Removal of these control inputs can be done by Tesla service or car owner and then the vehicle can be put into the Tesla robotaxi fleet.
A lot of knowledgeable people are saying this. Also, that Robotaxi is a ruse to prevent the Osborne effect on 3/Y.
As an Tesla investor this has been one of my personal bigggest test in faith in the mission and direction of the company. I trust the fsd direction and the AI robot narrative as a possibility. But now rather than being backed by the “growing car business “ we are gonna start going in the red and heavily investing in this path. It’s a challenge specially with macro economics. The market is technically not really rewarding pushing the limits it’s being more carfull than usual and it’s anticipating the big drop.
You need to do a valuation of RT vs M2. Mine shows robotaxi as a no brainer. M2 will not be the success you think because of competition from Chinese designs that will find their way in (built in the US or Mexico or Canada).
Evening guys
Evening, good Sir.
So, if this is really happening what are the initial use cases?
I’d like to hear discussions about that, for example in Manhattan. . .
*Success would have high visibility and bode well for wide international acceptance.
*The island is implementing congestion pricing south of 96th street.
*The maximum speed limit in this zone is 25 mph.
*The mayor is not just receptive, but promoting EV adoption.
*Is the FSD distributed ML model prepared for such an environment?-there may be more edge cases than highway driving.
If robotaxis are Tesla's next business, there will be no need to forcefully lower car manufacturing costs to the point of revolutionizing them. All you have to do is modify the existing Model 3 and install FSD. There is no need to build a new platform, so you can commercialize it much faster. I don't understand why they had to build on a new platform.
self docking for induction charging !
Shepherds will be Tesla employees, at least at the beginning.
Has Tesla/Elon thought about situations when there is a power outage and FSD data centres go down or the internet goes down…. What’s the contingencies being planned?
FSD is loaded on the vehicles hardware, does not require an internet connection
The FSD AI is installed on the car. It can operate in a tunnel with no signal.
As noted FSD operates on the car without need to communicate with a data center. But the dispatching will come from the data center so if the car loses the data link it can't get new rides or alert Tesla of any problem with the car (that needs to be passed onto the owner). In this case when the car starts a customer ride it can also pre-load a nav route to a safe parking place. Then if it loses the communication link it just drives to the "safe" place after dropping off the customer.
Finally, Datacenters have backup power generators and often power feeds from two different substations so power loss events are minimized. Also these type of systems are designed to forward all incoming transactions to a "backup" data center elsewhere on the internet that can provide seamless recovery of transactions when a Data center fails for any reason. So with two data centers, each can be the other's "backup".
Thank you guys, I didn't know.. it's good news.
If selling m2 gives Tesla $5k profit, but a robotaxi can bring in $25k/year for at least 5 years, that is a 25x difference in revenue.
That is hard to pass up.
I think James said 10k cars for Chicargo
They need car insurance, especially if a city bright by the border of another state
They are unstealable and hardly ever crash. Tesla will happily insure them with Tesla insurance.
Great discussion. "There's a lot to be done..." Amen to that. What is not generally appreciated is the scale needed to make RoboTaxi happen. RT is not Uber, RT is not rideshare. RT is a new form of transportation. A better way to think of it is "Personalized mass transport". Think about the "sherparding", or servicing of the cars. Zero in on charging alone. Each car in service will need to be charged multiple times daily. If we envision 500 robotaxis belonging to a service center (one every 5 square miles), 436 service centers would be needed in a metro area the size of DFW. That's 436 20-node supercharger ports, a total of 8,725, in DFW alone! Put in perspective - the US currently has 2,234 sites with 25k ports. Can Tesla scale it's supercharger mfg'ring business fast enough to support RT?
I would think Uber would like to transport folks around without worrying about the old employee/ independent contractor problems , not to mention they delivering , and other food and grocery .companies , and maybe even other deliveries like Amazon etc having no human employee drivers needed to pay and and worry about all those expenses versus a robot doing all 24/7 , no overtime pay no more vacations and work comp , disabilities etc.. imo that's huge maybe even hertz avis and other car related rental car cos will expand to this route !?
Case Example: Honda Resale Value…
I Leased a 2017, Honda Accord Sport. $0 Down $399 Month, $27,000 Buy.
Due to the safety, user dependability value and resale value of a Honda, I was hit and totaled the vehicle in an accident but received $3400 in my pocket due to value rise of the car.
The car was valued at $30,400, when I leased it at a $27,000 purchase price. ⛳️💰🧟♀️
total outlier case 😅
99.9% cars have their value depreciate quickly
@@maxs351 I know! 😁
That’s why Add Value to Tesla in the Marketplace is “Add Value.” ⛳️💰
The biggest hurdle is going to be convincing people to ride in a self-driven car. Even if the average FSD car has a better safety record than the average human-driven car, public perception is going to be hard to change. If the robotaxi model costs less to the consumer than human-driven taxis, that may get people to choose car service over car ownership, but it's going to be a long, uphill battle to sell the public.
The Robo taxis will need to have an insurance solution
If you believe that, the ordinary consumer of a Tesla could converted into a Robo taxi service and that means that the current fleet of cars are suitable for Robo taxi and we don’t need something with no steering wheel. Cruise has a steering wheel. They just don’t use it. The fleet is ready to go. You just need dispatching service and better FSD.
If the edge cases are becoming less and less and if you filter out all the non-edge cases, surely less and less compute power is required.
At this point I don't see the Roadster happening. No need to apply mfg resources to build a low volume halo car if Robotaxi is the focus.
Roadster is certainly off the menu. The real question is about model 2 and whether it is needed. A lot of investors seem fixated on it despite that is will be 5x less profit as compared to robotaxi. I think they need to just trust that they are not smarter than Elon. Armchair manager is an easy job.
Halo and pride. As long as the Roadster can be sold without a loss, then why not?
@@bobwallace9753 Opportunity costs. If the same human and factory resources can produce more profits producing something else, then that's where they will be deployed.
@@michaelnurse9089 Smart as Trump who's famous for stiffing contractors.
@@cascadiadesign
The major efforts will be in less expensive EVs. Especially robotaxis, IMO. But Elon is competitive. I think Tesla will spend very small portions of their overall expenditures on building EVs that outperform those of other companies.
I wonder if the Roadster might be built with gigacastings that could also be used for premium sedans and SUVs/crossovers which replace the S and X. Seems like it would come down to wheelbase width. The center section for each model could be unique.
Tesla will be able to add a steering wheel and other minor mods to the robo taxi production line if they want to change it over to model 2, IMO.
😮Don't you know that TESLA stops making MODEL2&REDWOOD ?
PLEASE STUDY MORE IF YOU LOVE TESLA.
@@MaLo55555 oh no! They're not going to race to the bottom! How tragic! Robotaxi is super useless compared to trying to make the cheapest EV, right?
Pipe dreams. 💨
NYC's new rules reqiure a SAFETY DRIVER BEHIND THE WHEEL. So Elon best not remove controls too soon ;)
If Tesla was to deploy 100 thousand Robotaxis in Chicago they would need 100 thousand indoor parking spaces for the cars. I've been to Chicago in the winter. During a blizzard all of the cars need to be indoors or they'll be frozen in place and it will be one hell of a job to recover them. Where will these parking garages come from? The problems are endless. What about charging? Remember the issues that Chicago had this last winter. Where and how will these Robotaxis Charge?
There will still be humans involved, to clean cars or remove them from snow. Or as a backup driver, maybe 1/100 cars, when they got stuck in outlier situation. Regarding weather, there will be days when operating safely is just not possible, but neither would be driving on those days!
Charging is easy to
solve, every roomba dust robot can dock itself on to a charger etc
@@maxs351 My question was on the days when operating safely due to weather is not possible just exactly where will they store the cars. Indoor parking spaces in a big city if they exist cost a lot of money.
They will more likely start in major cities in Texas and Florida. The cars will likely run non-stop in cold regions, internal heaters keeping the ice from building up.
@@michaelnurse9089 You obviously don't live in a city like Chicago. In a freezing cold blizzard the heaters would have to be on the outside of the car to keep ice and snow from building up. The cameras would also need to spray antifreeze and have wipers on them to keep them clean from the dirty slush on the roads. Obviously Texas and Florida would be easier.
It will take a few years to get the robotaxis built and the regulatory approval. There is time to work out the details of how this is going to work. It may quiet some of the investor rumblings if Elon comes out with some kind of roadmap showing what they plan. Not at a quarterly earnings call, perhaps at a special event, maybe the 8/8 day he already announced. I would have preferred this not to be an issue, but we don't live in a perfect world.
There is not enough detail into the cleaning and maintenance and charging.
If it is a regular owner car maybe the owner, but fast cleaning is difficult. So we don't have a car wash yet with interior cleaning for driverless.
Also, i have heard no comment on a significant increase in battery Reconditioning and replacement. It will be important because you will have a one or two year car needing it
And i see existing carsxwith Robotaxi capabilities same thing. In other words, a significant amount of battery care.
This is totally doable. I can easily see older model 3 improved with newer batteries
Now if Tesla owns they manage and pay for. But i am assuming this will increase as needed.
Let's assume a Model 3 goes to 300K. It's only 3-5 years old, in a no rust area. That in the planning will be expected. And for property taxes, more profitable for the owner. The car may be taxed at a 10K price but new battery goid to go. And Tesla will have access tolatest battery tech, new batteries are cheaper and soon to be a million miles.
Tesla needs to hire fleet managers from Uber and the like. Are there any job posts for this position!???
How about other players in robotaxis? Waymo, Cruise, Baidu, BYD, Mobileye …
You assume Tesla is the only one?
And how big is the market of Robottaxis?
In my particular case, I rarely used Uber because I have a car.
None of them have a chance. Tesla's training data lead is insurmountable and these others (excluding BYD) have no ability to mass produce BEVs profitably. Now, for a company like Mobileye, it would be SaaS only, so manufacturing is not their problem.
All dead or dead in the water. Waymo adds $200000 in equipment to their vehicles and they still don't work without remote drivers. China will get it working but Western nations will ban Chinese autopilot because of cameras being everywhere.
Tesla can undercut fair prices, because no driver needed. Will kill uber because of the fair prices. Waymo and Cruise are geofence and hardware that cost 250K Per car, can not be scaled affordability.
@@machoopichoo2
Few years ago, “game over 4680 battery “ and “game over $25,000 Tesla”
Are they now?
Now new theme emerges “ robotaxis” and only Tesla can do it.
Only Tesla can produce cheap ev?
Btw, if AI software is extrapolative, data is not that important because data is generated.
You need to drive the road twice, you already have all data. Subsequent drives are superfluous.
Tesla FSD is interpolative so data is primary, but prone to the edges ( new situations that are not yet in the data space)
@@Kukaboora Nobody said 4680 or $25K car were "game over." BYD can clearly make a cheaper car (albeit a death trap) than Tesla is willing to do. BYD and Tesla have been complimentary to each other and are clearly the leaders. It's a winner take most situation. Did the iPhone have a "game over" moment, no, but are they the leader, yes. You do know the Model Y is the world's best selling car of any type, right? Not "game over," but still leadership.
Explain how Tesla's approach to AI is that different than say Open AI (which Elon was a founder of). Extrapolative vs. interpolative? FSD is already extrapolating from real world driving to reading street signs and driving roundabouts it has never seen before.
Sure, FSD is particularly good in the Bay area because that's where most of the data is from and their engineers primarily reside. However, you clearly have no idea how neural nets train. Tesla has a billion miles under its belt and what does Waymo have, 10M lifetime. Tesla does that in a day, if not hours. Data "not important," so why do Open AI and Gemini scrape the internet for language? LLMs are generative, not extrapolative, from nothing. Tesla uses simulations, but this is not enough. Real world training is required for FSD to become a near AGI system. What is the path for others to achieve autonomy, genius...explain how they are going to do it other than just setting up an AI and all the rest just works out?
Lastly, great if others can (and they won't) catch Tesla in autonomy, but who, outside of BYD, can actually produce BEVs at scale and profitably to use their autonomous software? The entire industry caved on something as simple as charge plug form factor and adopted NACS. If all the others can't even get the plug right, how are they are going to figure out autonomy?
I don't know how existing Teslas (and future non-robotaxi Tesla cars) can operate as robotaxis with functioning steering wheels, brakes, and accelerators. How could the cars keep the passengers from taking over the automatic driving and creating havoc on the road? It seems to me that the only proper way for a robotaxi to function is to eliminate this possibility, so the idea that the millions of existing Tesla cars can be sent out to earn money for their owners is a pipe dream. I could be wrong but without steering-by-wire and braking-by-wire this doesn't seem doable.
see my reply to your other post below.
Robotaxi without sales of the Tesla Corrolla to the world is a mistake. Too much invested with too slow a payback.
Plus, FSD doesn’t work.
FSD V12 works great.
I really wonder about this FSD thing. FSD on it's own just does not work. Main reason is dirty cameras, darkness, weather and extreme busy and tight road conditions in Europe for example.They obviously know this. So i hope for them the new model has the required tools to fix all this.
Individuals wanting to use their personal vehicle as a robotaxi may run into problems with their insurance carriers.
A week ago, Tesla offered a free one-month trial period of FSD v12.3 to all Tesla owners in the US. Does anyone know how many owners tried it ?
This would be good information. Maybe they will reveal this during earnings on Tuesday.
I don't believe that existing Tesla cars, and future cars that are not robotaxis, will ever be able to go out and function as robotaxis as long as they have functioning steering wheels, brakes, and accelerators. What is going to prevent the passenger from taking over the car? Yes, Tesla servers could disable the car if it senses this has happened, but it still seems like a major flaw in the safe operation of the vehicle. Besides who is going to permit their expensive automobile to be operated as a taxi unless bought used specifically for this purpose.
At least a partial solution for existing cars would be to not allow a rider in the drivers seat and require seatbelts fastened on all occupied seats before the car moves. Then when a seatbelt unlocks and/or the interior camera detects an attempt to interfere with the controls the car, as quickly and safely as possible, pulls onto the shoulder and parks. Meanwhile Tesla alerts the owner of a problem and the owner can review the video, talk to the passenger (perhaps it's an emergency) and the owner can decide to alert the police to the car's location and starts the car lights flashing and sounds the horn. The car could tell the rider to refasten the seat belt as soon as it detects the unfasten event.
The driver door in taxi mode stays locked so a rider can't get in the driver seat from outside. So the only way a rider can take over is to crawl over the console. If this happens before the ride begins then the car stays parked and the owner is alerted. There is a window of time while the car is moving and the rider unfastens the belt and crawls into the driver seat before the car can safely pull onto the shoulder (consider the car on the freeway with no shoulder space or equivalent). Now the rider has control of the steering wheel at speed and the car can't stop without risking an accident. In that case Tesla alerts the owner as above and starts flashing the lights and horn and slows the car down gradually to 1/2 or 2/3 the speed limit so the police can stop the theft.
Also, fill the car in the back seat first so the passenger seat only gets occupied last. This makes it harder for a single rider to interfere with the controls. The owner could relax this restriction for riders needing the back seat for baggage.
Finally, Tesla could provide a replacement steering wheel for robotaxi operators with a locking clutch mechanism that stays unlocked during taxi operation so the steering wheel just spins freely. The throttle pedal could be disabled in software so the operator could use all 4/5 seats. But I think the breaks still have a mechanical linkage so if a rider was allowed in the driver seat they could still interfere with breaking.
A robotaxi fleet will require people and systems to inspect and clean vehicles, a system to send vehicles out for repairs , a place to park the vehicles as needed like during no drive snow storms, a place where the charge can be topped off, and a process system for maintenance and replacement, and a software system for tracking and monitoring all these fleet functions. What I've described is what rental car companies are already doing so I see Tesla buying the biggest, most widespread, most automated rental car company and replacing their ice fleet with Teslas. This will save Tesla years of having to hire for all those positions and building out all the parking/maintenance facilities and setting up all the fleet maintenance software and systems. Tesla could deliver 100K cars to inundate Chicago within one quarter of vehicle output. A ton of vehicle maintaining people, infrastructure, customer service people, claims/insurance people, and the like are going to be needed.
There are other ways to solve cleaning. e.g. contract with local car wash outfits and have the robo-taxies drive themselves to the car wash when the un-wiped part of the wind shield gets too dirty. Then have a car wash attendant give the interior a wipe down and vac. All scheduled automatically. Cars don't need cleaning every day. A car designed as a robo-taxi could be designed with extra cameras so it can alert a customer if they leave a package behind. Same system could detect someone getting sick in the car and send it to the car wash. Or a customer could alert Tesla that the car is dirty (not the best experience) and Tesla could dispatch another car and send the dirty one to the car wash.
For charging, Tesla needs to send the car to a storage lot with chargers and have an attendant plug them in or install inductive chargers (like the Apple watch or electric tooth brushes) so no human intervention is required. Or Tesla could start installing inductive chargers at super-charger locations and the robo-taxi's could go there. Tesla has been working on inductive charging so the Robo-taxi could have that support built in.
Tesla has service centers for repairs of customer cars so why not robo-taxi's too?
Clearly, any operator running multiple cars needs a place to park them (could be outside the city even) and handle stuff like window scraping, cleaning, charging, etc. Tesla won't be starting with 100,000 cars in any city to begin with. Today one can use the Tesla app to run the defroster to melt snow and ice on the windshield. If Tesla divides the city up into sectors like pieces of a pie then cars in any sector could drive to a parking lot on the perimeter of that sector slice. There are other ways to divide up the problem so it is solvable.
Tesla has very advanced internal software systems that run their production system and the business side. Musk started out in software and Tesla cars have by far the best integrated software systems of any vendor (autoline after hours did a piece on this). I'm highly confident Tesla can create their own taxi fleet management software since they already do that for their customer cars to manage service and software downloads automatically. Tesla needs to create the taxi ride request and dispatch software and these types of systems have existed for years, first in the airline business 50+ years ago, in the trucking industry, uber, etc.
The drawback of buying a rental car company is picking up a payroll of people some of which you don't need for robo-taxi's, like customer agents. Then integrating computer systems is always a problem along with integrating two corporate cultures. And that still doesn't get them the needed ride request and dispatch software. Lastly, does Tesla even want to be in the car rental business with a ton of ICE cars? Hertz profit margin is 6.57% and their operating margin is -15.5% (ttm).
@datamatters8 extremely thoughtful comment and worded respectfully, thank you.
The carwashes around Minnesota would need to have hired attendants to clean the inside of cars as all but the super expensive high end ones are self serve cleaning/vacuuming with automated washing equipment and the payment machines would need to be completely upgraded to accomodate autonomous cars and billing. Could be done but would require lots of collaboration, coordination, and contracts on a carwash owner by owner basis. Seems time consuming and daunting to set all of the aforementioned up. I'm sure Hertz has all those figured out, staffed, and systematized already. Your thoughts?
Why does nobody see that Tesla needs to develop their own maps and street view system. Many FSD trip failures are caused by obsolete map data furnished by Google. That is a key development for robotaxis to become fully functional.
Elon might be able to lower the prices of the current Tesla Y just by using the profits from robotaxi thus no stress with the model 2 in the end . FSD is 100% margins but in the end crushing the competing even more. Now
Imagine the price of Tesla cars if Tesla introduced 100-500k Tesla robotaxi. And introducing model 2 for different parts of the world that needs smaller cars
Starting manufacturing in Austin is already boutique vs Mexico 2 million odd.
It’s a whole new technology (many new tech, methods).
Making a small number of low price runabouts for private sale won’t be much income.
Can’t sell early units for $100,000.
Haven’t even got battery supply up & running (CATL line? Panasonic, LG 2170’s?)
Making private cars with controls, with forward facing seating, isn’t a help to profits, it’s just delay tactics.
That’s not balls-to-wall.
I think Tesla is going to double up existing factories. Tesla stockpiled a whole monsterload of batteries in Q1. LG revealed that. If you are sure then short the stock. I will stay invested for this as I think it taking TSLA past MSFT market cap.
@@michaelnurse9089 you misread.
I’m not shorting.
I’m saying story now is 100% the robotaxi.
Individual car ownership is a bad model, for consumer, for cities.
Production is Austin for now. Mexico hasn’t been started, Shanghai, Berlin, India, Indonesia aren’t moving.
(and soon enough AGI/ robots).
The mental gymnastics people go through to be in the tesla cult.
Off Topic, We need Tesla AI because so far the government and big business hoard AI. I need my robot computer to help me as defense for getting ripped off constantly, and propaganda. Jeannine
Didn't Hertz just got rid of their e.v. /Tesla fleet?
go through the game plan with only releasing model two. Manufacture cars, save nickels and dimes compete with everybody or totally dominate the car industry and make a ton more money. if model two sells slow down then make Robo taxis.
You get it. I get it. Elon gets it. The rest of the World think its nucking futs.
I believe that Optimus' first job will be to plug in & unplug Robotaxi from superchargers and then clean the interior while it's charging.
Remember when Homer Simpson was let design a car for Uncle Herb’s car company?: Cyber Truck
Remember when a over the top salesman sold a Monorail to the Simpsons town of Springfield?: elon musk
Since current Tesla EVs on the road are not steer by wire might it not be necessary to limit those EVs as robotaxis to licensed drivers? We don't want a seven year old behind the wheel.
At least a partial solution for existing cars would be to require seatbelts fastened on all occupied seats before the car moves. Then when a seatbelt unlocks and/or the interior camera detects an attempt to interfere with the controls the car, as quickly and safely as possible, pulls onto the shoulder and parks. Meanwhile Tesla alerts the owner of a problem and the owner can review the video, talk to the passenger (perhaps it's an emergency) and the owner can decide to alert the police to the car's location and starts the car lights flashing and sounds the horn. The car could tell the rider to refasten the seat belt as soon as it detects the unfasten event.
The driver door in taxi mode stays locked so a rider can't get in the driver seat from outside. So the only way a rider can take over is to crawl over the console. If this happens before the ride begins then the car stays parked and the owner is alerted. There is a window of time while the car is moving and the rider unfastens the belt and crawls into the driver seat before the car can safely pull onto the shoulder (consider the car on the freeway with no shoulder space or equivalent). Now the rider has control of the steering wheel at speed and the car can't stop without risking an accident. In that case Tesla alerts the owner as above and starts flashing the lights and horn and slows the car down gradually to 1/2 or 2/3 the speed limit so the police can stop the theft.
Also, fill the car in the back seat first so the passenger seat only gets occupied last. This makes it harder for a single rider to interfere with the controls. The owner could relax this restriction for riders needing the back seat for baggage.
Finally, Tesla could provide a replacement steering wheel for robotaxi operators with a locking clutch mechanism that stays unlocked during taxi operation so the steering wheel just spins freely. The throttle pedal could be disabled in software so the operator could use all 4/5 seats. But I think the breaks still have a mechanical linkage so if a rider was allowed in the driver seat they could still interfere with breaking.
yeah, compact is delayed, lets see
TESLA stopped making MODEL２&REDWOOD.
Why NOBODY knows ??
people like driving
Not all the time, heavy or city traffic, or long distance are bothersome to most. And most importantly people get distracted and cause accidents. Safety is a main issue.
People also like riding horses
100k cars needed only for Chicago? S a y. W H A T?
bad news Wallstreet will drive it up to dump the stock!
One Shepherd could cry wolf lol
Sounds like 2-4 years to get through the sticky part. I’m ok with that.
Initially the 25k car is going to sell faster then they can produce them. They will sell to lower and lower middle income people. If sold to them correctly, as a way to pay for itself and put extra dollars in their pocket, they will use them as robotaxis. I think that will be the group willing to share their car with strangers, if the $ really pay for the car and give them extra income they need badly. Middle class and up are not going to be willing to share their car. They will take dedicated robo taxis, but will not share their car.
the present models doors don't shut by themselves
Have the car play a recorded message to close the doors when the trip ends. They car knows the riders are getting out via the seat belt opening as well as the cabin camera. If the rider does not close the door after 10 or 15 seconds play the message again. For a rider that wants to screw with the car owner then the car can just periodically just play the message asking for help and sound the horn before each message playback and someone else can close the door. This should solve the vast majority of cases. Otherwise the car (via tesla) alerts the owner to the problem and the owner has to go retrieve the car. But since the owner has the rider's id and credit details the owner can charge the driver a fee for the recovery and record the event on the rider's account so other robo-taxi operators can exclude the rider in the future as desired. This doesn't seem to be a big issue to me.
I hope that Tesla doesn’t operate a robotaxi network. I’d like to see fleet sales only with recurring revenue for software
I disagree with Larry on the length of time it will take for RoboTaxi to spool up.
Those things he is concerned about are easily and quickly solved
On Hertz Fleet - First Hertz Tesla fleet was just dissolved - total disaster Hertz entirely to blame - not Tesla - Tesla has to manage deployment - not just do sales of Robo - Tesla should own - especially all data collection - & lease -franchise? - commercial / ups / fedx / delivery / Pepsi ? Sysco is real opportunity other than urban taxi - deployable now
Why would a company use their smallest vehicle as a taxi. The Y or X would be a much better taxi. I have a m3 (love it), but way to small for a taxi.
Why do we think Elon, at his current age and in Today's world, is the same hungry Elon that he was 8 or 10 years ago?
There is no evidence that this is not the case.
He hasn’t landed a colony on Mars yet
What difference does it make? Musk throwing tantrums around his staff does not make robotaxis work.
Elon is only 52 years old. That's when people are at their best. More experienced and wiser than the youth and higher energy than the elderly.
Musk has plenty of money to stop working any time. But he has life goals that matter to him and that is what drives him. Not more money.
Robotaxi and Compact car are same program....On 08/08 Tesla will probably show the robotaxi prototype which will be the same as compact car....With and without sw....surprise will be huge....
The only difference is DRIVE BY WIRE!…..easy peasy
Hope you are right
25k robo taxi for personal use. Maybe a 2 seater.
general supervised fsd is more important .... use Model 2 car sales to help fund/stabilize ... and support the 10 year transition to robotaxis that may or may not be possible
Its probably just an OTA update that you can turn your tesla into a driverless uber 😅
You will have to sign a contract with Tesla because Robotaxi is going to cost 10x what FSD costs.
@michaelnurse9089 yeah it'll be like the FSD subscription