I have it in writing from Tesla that the post-delivery FSD price that was advertised at the time of purchase is what an owner will pay, no matter the currently advertised FSD upgrade price. The web site charges current price and any difference due to the owner, based upon the advertised price at time of purchase, is credited back to the purchaser's credit card in some number of weeks.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
BUY FSD TODAY!
- Thread starter Ezri10
- Start date
I have it in writing from Tesla that the post-delivery FSD price that was advertised at the time of purchase is what an owner will pay, no matter the currently advertised FSD upgrade price. The web site charges current price and any difference due to the owner, based upon the advertised price at time of purchase, is credited back to the purchaser's credit card in some number of weeks.
One thought I had relevant to this thread is how Tesla will handle the influx of service requests for HW3. While it will be less than an hour service, survey that will hammer service centres.
Perhaps Tesla will send out invites in waves, starting with people who have had FSD the longest. This will surely be welcome for those who’ve been waiting a while and justify a long wait service appointment wait for those that haven’t paid for FSD yet.
Just a thought in case getting FSD features as soon as they become available to you.
Perhaps Tesla will send out invites in waves, starting with people who have had FSD the longest. This will surely be welcome for those who’ve been waiting a while and justify a long wait service appointment wait for those that haven’t paid for FSD yet.
Just a thought in case getting FSD features as soon as they become available to you.
Knightshade
Well-Known Member
If they're as easy to swap as suggested they'll probably have rangers go to you to swap em.
For sure. I can see rangers too. Regardless, there will be an influx of service requests and presumably bottlenecks.
In 6 months there will be how many HW2 cars on the road? Upwards of 400k? How many of those will have bought FSD once HW3 is available?
I didn’t read all the comments as there are many.
I support the OP in that we are getting closer. Here is my take:
Tesla announces new hardware is free upgrade for people that bought FSD. Now everyone adds FSD to their order, me included, and they realize that everyone will get FSD just for the updated, currently unneeded hardware. This is where the confusion is. They don’t want to give away all these free upgraded computers to everyone that purchased FSD, also they need a hard number to plan on how many upgrades will be free so they stop taking orders for it now.
I feel that they ultimately stopped selling FSD not because they’re not close but because they ARE close. Think about it, they need 200 employees to test it out for more data, this would be like a closed beta test before opening it up to an open beta when the new hardware comes out.
This is great news and I’m glad I bought FSD.
I hope you are right. If they are close to true FSD it would be wonderful and I'd be happy to pay the increased cost of not having pre-ordered.
BUT...
Look at where EAP is now. Look at where all the other car companies are with their driver-assist packages. And look at where Waymo, the company furthest ahead with FSD research, is. It's just not rational to expect a breakthrough that big in the near term. Tesla has not even begun dealing with traffic light recognition or small-obstacle recognition.
Tesla probably is very close to a bifurcation of the "FSD" package from the EAP package, where some new feature will be introduced to the former that is not given to the latter. Speculation is on telling you what lane to be in on the freeway (something Google Maps does now) or maybe actually changing lanes without your having to take any action to tell it to. Some people think maybe Level 3 (no nag) on the freeway. But true full self-driving is a decade away.
Bill Price
Active Member
Prognostication at it's finest. If only the rest of our crystal balls were not so translucent. As Butch Cassidy said to Sundance, "Boy, I got vision and the rest of the world wears bifocals."
Tesla is working on it, but it's not ready for release.
Yes. Tesla must have something amazing up their sleeve to fulfill their promise in the foreseeable future. OR my guess is that the maturity in FSD is reflected in EAP since EAP is a subset of the whole package which is FSD. FSD needs EAP to handle turns, start/stop, lane keeping etc, and FSD is the overarching logics and more neural networks etc for intersections and all that stuff.
I strongly doubt the FSD is way ahead on a completely different source tree. EAP is the demo/posterchild for their self driving capabilities, so if they had something more stable right now, why not just release it into EAP?
We know that the current AP V2.0/2.5 hardware limits the size/complexity of the neural network that Tesla can run on its current vehicles. From Tesla presentations, we know that Tesla has to make tough decisions about what to include in the neural network (i.e., the types of situations to handle) due to V2.0/2.5 hardware limitations, and that optimization is critical. Tesla has said that the V3 hardware will support a larger / more complex neural network.
As I see it, there will be two neural networks — one for V2.0/2.5 hardware and one for V3 hardware. The two will be similar except the NN for V3 hardware will handle 6-10x as many situations, sufficient for some level of FSD (be it 3, 4, or 5) as well as improved AP.
Last edited:
Knowing a little about autonomy, and then having owned my Model 3 for 2 months, I just don’t understand how anyone can think that FSD is “close”.
Now, that’s not to say that Tesla won’t make some of the Autopilot features exclusive to those who own the FSD upgrade, but real, FULL autonomy is truly 7-10 years away. There are far FAR too many places where roads and infrastructure simply isn’t ready even if Tesla was. I’m incredibly glad I didn’t spend the extra money. My next car may have it, or even the one after that.
Now, that’s not to say that Tesla won’t make some of the Autopilot features exclusive to those who own the FSD upgrade, but real, FULL autonomy is truly 7-10 years away. There are far FAR too many places where roads and infrastructure simply isn’t ready even if Tesla was. I’m incredibly glad I didn’t spend the extra money. My next car may have it, or even the one after that.
Yes it will be more powerful, but again it uses many more cameras. It's more about how it behaves which doesn't require more computational power. This is NOT a hardware power issue. LKAS systems uses a small computer that's nowhere near AP 2.5 hardware. I am talking about stuff like it taking bad decisions on a higher level:
- Decides to follow another lane right into a divider
- Positions itself in the middle when there's a merging field, instead of just holding it's original position in the line that's always there
- Not caring about other cars merging in front
- Teenage-driving-braking when cars move in front, heading towards a traffic jam
- Ghost braking
I would be amazed if they nailed this. Otherwise you MUST have what he calls "Software 1.0" on top of deep learning black boxes. and I feel that it lacks a lot of this situational, overarching logic. They have enough trouble training their NNs.
The L1, L2,L3,L4 and L5 levels are just broad categories for ways to describe car automation. They are not absolute technical levels or even human skill levels. (Some days I’m pretty sure I’m barely qualified to drive in the rural neighborhood.)
What I hope HW3 implements is handsfree driving on the limited access expressways. Yes, it’s harder to drive in the city or on unmarked roads without cellphone reception, but just working on the interstates would allow me to travel long distances.
So I look forward to ( basically just expressway driving):
Limited access handsfree driving
Handsfree driving in construction zones
Reading speed limit signs ( permanent and temporary )
Drive on Navigation without human intervention
Dealing with emergency vehicles on expressways, stopped or otherwise
That should be possible in 2019.
What I hope HW3 implements is handsfree driving on the limited access expressways. Yes, it’s harder to drive in the city or on unmarked roads without cellphone reception, but just working on the interstates would allow me to travel long distances.
So I look forward to ( basically just expressway driving):
Limited access handsfree driving
Handsfree driving in construction zones
Reading speed limit signs ( permanent and temporary )
Drive on Navigation without human intervention
Dealing with emergency vehicles on expressways, stopped or otherwise
That should be possible in 2019.
They are releasing better firmware versions all the time. The EAP in my car is more stable now than when I got it. They claim that the next big step needs a more powerful computer, which they're getting ready to introduce. But I'll be flabbergasted (and delighted) if they're ready for L3 (eyes off the road) on the freeway in less than a couple of years.
I expect that even features like stoplight recognition and speed-limit sign recognition will continue to be Level 2 (full driver attention required) for several years to come. Same for Drive on NAV even if it's just on the freeway. I also expect to see some confusing mixes in five years from now, when EAP is level 3 (eyes off the road) on the freeway but "FSD features" like Drive on NAV in the city remain Level 2 (full driver attention required).
Tesla is one of the leaders in this field (though I think Waymo is ahead) and Tesla's cars have the best implementation in consumer-available cars to date, and driving on EAP today is utterly amazing and mind-boggling. But the challenges ahead before we reach even just Level 3 in the city are enormous, and while I firmly believe that Level 5 (driver optional) will come, I am not at all certain that it will come in my lifetime.
Note that even Level 3 on the freeway requires small-obstacle detection and recognition at better-than-human capability. I'm not yet comfortable using EAP when there are traffic cones on my side of the painted line.
I think it unlikely that Tesla will start giving away EAP for free any time soon. However, the history of safety features in cars is that at first they are optional extras and eventually they become standard. Seat belts, anti-lock brakes, air bags. But as long as EAP is a level 2 feature (driver must remain fully alert) I don't think it will be included in the base car.
Some day, but not for a while. Certainly not while only the top luxury cars, and Tesla, have it.
Some day all cars will be required to be fully self-driving. Maybe in twenty to thirty years.
Do it if you believe that the package will include features you think are worth the cost, within a time frame acceptable to you. Or if you are willing to gamble that it will. I am not a gambler. I once bet one dollar on a roulette wheel, didn't feel any thrill, and walked away. And I don't think the FSD package will have anything Earth shattering in it within the next 3 or 4 years. And I can afford to pay the higher price if I'm wrong.
My advice is no. Clearly, many people here would advice yes.
If you love new features, having the latest and greatest and take part of the beginning of self driving cars, sure go ahead!
If you can handle to miss out on some features, and can wait for full self driving when it's actually there (for whatever price that will be), then no.
Similar threads
- Question
