Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla Raising Price of FSD to $15,000, AI Director States "We Can Build a Car That Never Crashes"

This site may earn commission on affiliate links.
After FSD Beta 10.69 successfully rolled out to Tesla owners last night, Elon Musk tweeted that once it is released to all FSD Beta participants, the price of the FSD package will be raised from $12,000 to $15,000 on September 5th. The current price will be honored for any orders placed before September 5th.


Another price increase seems like a reassurance that FSD will come to fruition soon. Tesla's Autopilot Software director Ashok Elluswamy shares additional information regarding the progress Tesla has made in FSD Beta 10.69: a huge change introduced in this update are "Occupancy Networks," which Elluswamy states are "our approach to solve general obstacle detection and using it to enable sophisticated collision avoidance." He claims that these occupancy networks are so important in solving self-driving that "when we execute on this plan we can build a car that literally never crashes."



Tesla enthusiast @WholeMarsBlog on Twitter shares their first 35-minute drive with FSD Beta 10.69, highlighting the significant changes Tesla has made in this update.


FSD Beta tester Chuck Cook @Chazman (who also snuck into the v10.69 release notes) also shared a video testing how FSD Beta 10.69 handles turns. "It wasn't very smooth, but it was adjusting its course so it can continue without a full stop," referring to the unprotected left turn his car had just executed.


Screen Shot 2022-08-21 at 11.50.25 AM.png

Credit: Chuck Cook on YouTube

We're hoping all the changes that need to be fixed by 10.69.2 will provide more reasoning for Tesla's sudden price increase in FSD Beta 10.69. We can only hope this means Tesla is seriously getting closer to solving the self-driving problem, and we will certainly see how the changes allowing video and map data to simultaneously help FSD Beta learn to drive will affect not only how it drives now, but in the near future.
 
Over a decade? Been a decade since the S was first released; I don't recall any conversation about Autopilot until the A&D announcements in August 2014. That's 8 years, not over a decade. Did I miss anything else?
Right and Beta just started 2 years ago in October. It's made insane progress in that time. Even in the year that I've been a tester, it's made significant improvement. With 10.3, the car was maybe driving 50% of the time, now with 10.12.2, it drives for me 90-95% of the time with out the need for intervention.

All of that said, to get to 100% of the time and full level 5 we are years away and we may never completely get to that point with the current hardware.
 
Right and Beta just started 2 years ago in October. It's made insane progress in that time. Even in the year that I've been a tester, it's made significant improvement. With 10.3, the car was maybe driving 50% of the time, now with 10.12.2, it drives for me 90-95% of the time with out the need for intervention.

All of that said, to get to 100% of the time and full level 5 we are years away and we may never completely get to that point with the current hardware.
Years away? i think you may be mistaken. The CEO has repeatedly said FSD WILL be fully completed by 12/31/2022, which includes L5, as also stated by the CEO.
 
  • Funny
Reactions: boonedocks
Years away? i think you may be mistaken. The CEO has repeatedly said FSD WILL be fully completed by 12/31/2022, which includes L5, as also stated by the CEO.
Lol

"Fully completed" FSD in its current form = wide release to the fleet of people who have bought FSD, and then Tesla can realize the revenue on the accounting side of things. It'll still have a (Beta) tag and it'll still be Level 2, it may or may not still have a Safety Score requirement.

What happens after that is anyone's guess, hopefully it doesn't then sit for years while they break out into some new development branch trying to achieve Level 3+

This is exactly the path laid out to the California DMV way back when
 
Lol

"Fully completed" FSD in its current form = wide release to the fleet of people who have bought FSD, and then Tesla can realize the revenue on the accounting side of things. It'll still have a (Beta) tag and it'll still be Level 2, it may or may not still have a Safety Score requirement.

What happens after that is anyone's guess, hopefully it doesn't then sit for years while they break out into some new development branch trying to achieve Level 3+

This is exactly the path laid out to the California DMV way back when
Wait..so you are saying there wont be Robotaxi service by 12/31/2022? No ability to summon the car from NYC to LA with no human input by 12/31/2022?

 
Hypothetical - suppose he does deliver level 5 some day. How many people would then get dropped off at the office while your car Uber's all day?

Zero. How do you keep your car from being stolen and chopped? It just gets easier for the thief - summon the car to a dark, remote pickup location, disable WiFi, and off you go. If kids can steal a Kia/Hyundai in under 30 seconds, they will love Tesla in assisting in their endeavor.

No?
 
Hypothetical - suppose he does deliver level 5 some day. How many people would then get dropped off at the office while your car Uber's all day?

Zero. How do you keep your car from being stolen and chopped? It just gets easier for the thief - summon the car to a dark, remote pickup location, disable WiFi, and off you go. If kids can steal a Kia/Hyundai in under 30 seconds, they will love Tesla in assisting in their endeavor.

No?
That's a weird hypothetical and we aren't even close to that point where Tesla or anyone would need to create safeguards to prevent that.
 
There are many reasons consumer vehicles as Robotaxis would be insanely complex if not impossible to pull off, unless we think about a fully socialized system there nobody actually cares about their own vehicle — but we’re talking about people’s own vehicles, so that doesn’t make much sense.

Lets say you work from 9-5, get to work and send your vehicle off to drive around. What if someone gets in at 4pm and plugs in a 2hr long milk run? You would need to build in certain timeframes and buffers around them, and then users would need to be matched based on those criteria but that seems to rigid for reality.

What if you let your car abandon you at work and then have a family emergency while your vehicle is in the middle of a trip? You need to be okay with working around that, and another ride needs to be immediately available.

Vandalism, theft, and other such issues are a given.

There are so many logistical things that would need ironing out, but they’re all solved instantly by Robotaxis being a third-party service.
 
What if you let your car abandon you at work and then have a family emergency while your vehicle is in the middle of a trip? You need to be okay with working around that, and another ride needs to be immediately available.
In such case, I expect you will robotaxi home in someone else's Tesla? Since the streets will be clogged with out of work Tesla robotaxis, getting a ride should be pretty easy!

Seriously though, the idea of the personal robotaxi making you a fortune is a pipe dream. Current Teslas are not 'hardened' for public transportation - especially not unattended transportation.

Robotaxis, in general, are very vulnerable. How long before out of work Uber/taxi drivers learn that a piece of duct tape on the B-pillar stops your robotaxi cold until you, or some other technician, goes to its location to take the tape off?
 
Robotaxis, in general, are very vulnerable. How long before out of work Uber/taxi drivers learn that a piece of duct tape on the B-pillar stops your robotaxi cold until you, or some other technician, goes to its location to take the tape off?
Uber drivers are vulnerable too: ‘Violence guarantees success’: how Uber exploited taxi protests
Over time, Uber drivers were attacked in dozens of countries and even murdered by taxi drivers in South Africa, Brazil, Colombia and Mexico.
 
I'll again assert that raising the FSD price from $12k to $15k should have almost no impact on Tesla bottom line. The price would reasonably result in slightly lower adoption vs never getting it or renting it for $300/mo when needed. I don't think enough people could be "scared" into buying-now-before-it-goes-up-more to have any impact.

This is IMO pure ego - Tesla convinced to the core that they have the tech almost ready for people sending their personal car out on it's own as a taxi service, which simply does not look realistic any time soon.
While I think monthly rental is now the best option, the problem is deciding “when needed.” Having the Beta on both cars I want it daily not just when I take that once every few months lengthy trip and I suspect that is the case with most people. The extra set of eyes and fast reflexes makes FSD much more than just a tool to make long drives more relaxing. In spite of its many current foibles, I use FSD every time that I drive.

So what can I do about it? Simple. Hope neither of my cars gets totaled and drive them both into the ground. My cars are both old enough that I didn't have AP included in the original price. It cost $5000 for one car and $7000 for the other one to bring them up to FSD. It would be crazy to throw away that $12,000 expenditure and then pay another $30,000 for FSD on two new cars.

I can only conclude Tesla doesn’t care if those of us with “old” cars become repeat purchasers.
 
"Vandalism, theft, and other such issues are a given".

I believe we still have a long way on the technical issues before we have to worry about thwarting the criminal. Rental car companies have done a pretty good job of mitigating their risks in this regard so compared to the technical side, this is on the minimal side of things.
 
  • Like
Reactions: 2101Guy