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

Wait on adding AP2 and FSD or you will be sorry...maybe

This site may earn commission on affiliate links.

azred

Active Member
Apr 12, 2016
2,354
3,611
Chandler, AZ
It seems highly unlikely Tesla will charge a steep $8000 at purchase or even steeper $10,000 post-purchase to flip the software switch for the "affordable" Model 3, right? Therefore it makes sense that you should wait until at least July to see if the exact same software will also be cheaper for Model S and X. We have seen all of the angry posts from people who believe they were mistreated and misled by missing the battery-related price cut a couple days ago, so be smart and wait if you are getting ready to add AP2 or FSD.
 
I would expect Tesla to greatly increase the price of the FSD package once it starts offering meaningful capability. Sure those who bought in early are kickstarting Elon's self-driving system, but I think there's a bit of an early adopter discount baked in :)
 
I've been thinking the same thing. It will be interesting to see how the pricing works out.
The tricky decision is for those buying a new X or S before July. Do I not pay the $8000 now counting on a big price cut that may not happen? Personally, I think it's a big enough deal that I would hold off on buying a new X or S until the Model 3 configurator opens in July, at a minimum.
 
  • Like
Reactions: davidc18
I would expect Tesla to greatly increase the price of the FSD package once it starts offering meaningful capability. Sure those who bought in early are kickstarting Elon's self-driving system, but I think there's a bit of an early adopter discount baked in :)
Really hard to say. It can go either way. It can go up like AP1 did (to better "align" with the value of the system). Or it can also go down as the larger volume amortizes the fixed costs better.
 
I would expect Tesla to greatly increase the price of the FSD package once it starts offering meaningful capability. Sure those who bought in early are kickstarting Elon's self-driving system, but I think there's a bit of an early adopter discount baked in :)
To the contrary, I think Elon has made it clear there is an early adopter penalty baked into S and X. We have seen further evidence of that in the events of the past week. While it may always be the case that Tesla will try to gouge the rich folks who are buying the $100-150k models, I think it is highly unlikely that will be the case for Model 3 and the "middle class" versions of Model S.
 
  • Like
Reactions: Ulmo
I bought EAP & FSD when I got my car in December. I must say I'm disappointed at the pace of the roll out. Some may say that they have made great progress. And that is true now that we know they started from scratch after MobilEye. However, from my expectations in December, based on what we were told at the time, it has been disappointing.

I look at the video Tesla released for self-drive (on the streets to the office) and I wonder why is I cannot get anything close to that? I look at other car vidoes (Bolt for example Watch GM’s self-driving Chevy Bolt EV prototype slow down for a raccoon [Video]) and ask myself why can I not get anything close to that?

What is the answer? Are these videos so staged that they they are not even close to current capability?

It seems like Tesla puts out releases so often that we are getting their latest and greatest. If that is true then their automation video was fully staged. I don't know how else to interpret that.

All the others seems to take a totally different approach. They want their driving automation to be fully baked before release.
 
Last edited:
They may or may not be staged, but they are videos of an engineering trial that goes well. Think about the space program and all the videos of rockets exploding during the Mercury program before we successfully cracked the code. Occasionally a launch went well, but then another problem would crop up and they would have more rockets blow up. The videos you've seen are videos of a successful launch, but there are still enough problems and failures that the system is not ready for release to the users and is still in testing... What is unusual is that typically these kind of videos don't get released this early in the research and development cycle (especially in the auto industry) but Tesla is disrupting that approach by showing people what is possible....
 
  • Like
Reactions: Ulmo
What is the answer? Are these videos so staged that they they are not even close to current capability?

The demos are all totally staged, but it is disappointing that Tesla won't show us anything to raise spirits. :) Today's staged GM demos appear to be the most impressive, but they also didn't show any camera views with object detection to convince any engineers that they are doing any more than following a rigidly pre-programmed course with staged pedestrian traffic. The last Tesla demo was a standard Mountain View joke with off-the-shelf 2016-era demo ware. Every player has those.

All we know for sure is that Tesla isn't doing any FSD miles in CA because they have to disclose progress and disconnect rates, and they only did a tiny number of miles with LOTS of human takeovers months ago. We can only hope that they are at least working on L3 and L4 features and testing somewhere, but for now, there is zero to go on. It doesn't particularly bother me that they are not doing FSD demos, per se, because that isn't essential in building the technology. They could be doing LOTS of miles that don't count as FSD but are providing training data and validation.

Objectively, it pretty much looks like nobody is ahead with a commercially viable solution and all the high-end cars and EVs will probably offer similar capabilities somewhere around 2020-2022, if we're lucky.
 
I look at the video Tesla released for self-drive (on the streets to the office) and I wonder why is I cannot get anything close to that?

Because that was just one successful demo. I bought around the same time(Dec 31 delivery) and it was made pretty clear that at absolute earliest, we'd be getting FSD at the end of 2017. I'd also note that that GM video is *also* from one successful demo. They also haven't released anything close to FSD.
 
Did GM stage the raccoon? If they did I'm impressed. if not that is one lucky raccoon!

I'm sure for that one raccoon there are many many more flattened raccoons. Let's assume they were plastic decoy raccoons.


But seriously, a manufacturer-provided demo video is demoware no matter how you slice it. If anyone wants to live-stream their self driving footage, I will pay attention to it. Other than that, I expect them all to be curated to some degree — obviously some more than others.

If you live in Palo Alto and have followed a Waymo panda car around, I think you would not walk away thinking it's super impressive.
 
It seems highly unlikely Tesla will charge a steep $8000 at purchase or even steeper $10,000 post-purchase to flip the software switch for the "affordable" Model 3, right? Therefore it makes sense that you should wait until at least July to see if the exact same software will also be cheaper for Model S and X. We have seen all of the angry posts from people who believe they were mistreated and misled by missing the battery-related price cut a couple days ago, so be smart and wait if you are getting ready to add AP2 or FSD.

I heard the hardware alone from Nvidia is roughly $3000. Autopilot is a huge must have "optional" feature. I doubt they'll reduce price on it, simply because they don't need to. People will buy it at time of purchase or at some point in the vehicles life when AP gets fully implemented
 
The demos are all totally staged, but it is disappointing that Tesla won't show us anything to raise spirits. :) Today's staged GM demos appear to be the most impressive, but they also didn't show any camera views with object detection to convince any engineers that they are doing any more than following a rigidly pre-programmed course with staged pedestrian traffic. The last Tesla demo was a standard Mountain View joke with off-the-shelf 2016-era demo ware. Every player has those.

All we know for sure is that Tesla isn't doing any FSD miles in CA because they have to disclose progress and disconnect rates, and they only did a tiny number of miles with LOTS of human takeovers months ago. We can only hope that they are at least working on L3 and L4 features and testing somewhere, but for now, there is zero to go on. It doesn't particularly bother me that they are not doing FSD demos, per se, because that isn't essential in building the technology. They could be doing LOTS of miles that don't count as FSD but are providing training data and validation.

Objectively, it pretty much looks like nobody is ahead with a commercially viable solution and all the high-end cars and EVs will probably offer similar capabilities somewhere around 2020-2022, if we're lucky.

Also what is the price of the equipment being used? If GM achieved that with 1 million dollars worth of lidar and sensors then really who cares.
 
  • Like
Reactions: chillaban
Did GM stage the raccoon? If they did I'm impressed. if not that is one lucky raccoon!

In some sense, of course it could be staged. I'm not saying that the car didn't "recognize" the raccoon and do a stutter-step in speed, but for sure the raccoon could have been released at an opportune time or they could simply drive until that raccoon crosses the same exact street at the same exact time every night. Raccoons follow the same exact route every night. In my subdivision in Illinois, the same raccoons emerged from a sewer drain, crossed the road, wore the same path in the grass in my neighbors' yards, and then wore the same path in my yard, played in the same birdbath and climbed onto my roof the same way every single night.

Programming in a raccoon neural net and hoping for a crossing is not a general solution.

But, it is a damn good demo!
 
  • Like
Reactions: Ulmo
The tricky decision is for those buying a new X or S before July. Do I not pay the $8000 now counting on a big price cut that may not happen? Personally, I think it's a big enough deal that I would hold off on buying a new X or S until the Model 3 configurator opens in July, at a minimum.

If you already have a Tesla but were looking at upgrading, I'd be waiting until near the end of the year to buy an S/X - until after Xavier ships and AP presumably transitions to that (with or without new sensors,) maybe until 2170 cell packs come out (those might not be out until sometime next year,) and until Tesla makes changes to the interior and pricing structure in the wake of 3 mass production.

OTOH, any modern Tesla is so much better to drive than other cars that I have trouble suggesting folks coming to Tesla from ICEs should hold off buying on. Maybe it's CPO and inventory car time...
 
  • Like
Reactions: Ulmo
I'm not surprised that this thread with a narrow focus has veered off into numerous directions. That's what happens to every thread. I just want to emphasize that while I felt bad for people who believe they were blindsided by the huge price cut in 60-75 upgrade and the related new car price cuts, it was totally foreseeable. The only surprise in my view is that it happened so fast given that the upgrade cost was an astonishing $9000 a mere three months ago.

Obviously I am just guessing -- even Elon might not know the answer today! -- but I think it is equally foreseeable that the price of AP2 and FSD will be much less for Model 3 -- and the price for X and S will follow suit. There are many ways to keep S and X prices stratospheric without charging more for the cost of identical software. It makes no sense.
 
I made the following comments after I watched it at Electrek:

If this is what it seems to be, then this is a good first start. It is too wimpy about yellow lights and right hand turns and is poor at avoiding possible left hand turners, and would not be acceptable to passengers or other drivers if it elongated trips due to those three issues, and it looks like it doesn't have the best pot hole avoidance capability, but other than that, it showed abundant competence about all other issues. I applaud the work on this so far, and if they can bring to bear the proper psychological well being to approach those four issues, then this may become a successful product.
It seems highly unlikely Tesla will charge a steep $8000 at purchase or even steeper $10,000 post-purchase to flip the software switch for the "affordable" Model 3, right? Therefore it makes sense that you should wait until at least July to see if the exact same software will also be cheaper for Model S and X. We have seen all of the angry posts from people who believe they were mistreated and misled by missing the battery-related price cut a couple days ago, so be smart and wait if you are getting ready to add AP2 or FSD.
My strategy is simple; in order:
  1. Wait until the product meets my specifications for the listed price. This will be when:
    1. It drives better than me:
      1. Safer
      2. Faster
        1. Keep in mind faster takes into consideration my life speed; this can include amount of time I can do other useful tasks, whether I have other useful tasks in that environment, and the relative time quality of those tasks compared to driving. (For instance, I may run out of quality tasks that I can do in the car-driving-itself environment, thus not making the opportunity to do other things as useful as it could be if I did not run out of quality tasks to do. So far, I already have a large book list saved up, but I don't know if that will be sufficient.)
    2. It saves me enough life cost compared to its feature cost.
    3. It meets #2 below.
  2. Wait until I have the listed price in budget.
  3. Verify both #1 and #2 are still true simultaneously.
  4. Buy it.
I'm not getting it ten sextillionths of a yoctosecond before that.

Many other posters posted the same points I have:
  • EAP is useful now; consider the cost-benefit for you.
  • Consider the donation effect of giving extra money for the ramp up and development of Tesla projects.
  • FSD is not useful now; consider the cost-benefit for you to help support Tesla develop it with your money that you may never get back directly.
  • EAP, FSD, one or both, may or may not start to compete with Model 3 when it comes out, and have price drops. This could be due to lowering of cost of those options in competition (within and without Tesla), or due to the exclusivity factor in the cost of upper-range vehicles (higher cost S & X).
  • There is no guarantee anything in particular will happen in the future, but we can see general trends. AP1 never got to where Tesla said it would get to; I'm not assuming AP2 will either. Both are improving. Both have their use cases. Both will change in price over time, up and down. Both sometimes get worse.
  • Most of us who look carefully at this issue think newer hardware will come out to address some key issues: vantage points, variety of sensor capabilities (such as night vision, animal and object detection through difficult terrain, etc. (xray and infrared come to mind as some techniques)), and computational power. The one thing I'm not worried about is computational power since that's upgradable, but they might rescind (never offer) the ability to upgrade it or not offer great upgrade options.
There's a positive effect both from getting it now and from waiting, and how those two balance for the particular person will be important. I think many people here can justify the expense of getting one or more of the assistance options, and many here cannot.
I'm not surprised that this thread with a narrow focus has veered off into numerous directions. That's what happens to every thread.
That's what's supposed to happen to every thread. Thoughts grow to new thoughts. The fault of the situation is the non-tree based flat linear message system that TMC selected. It is simply due to inadequate prior examples in the Internet and the experience of most people who use it; it is a fairly typical situation.

One example of a partial tree system is Reddit, however, it has the problem that it still requires you to select a topic ("subreddit") for each original posting, and those original postings cannot be organized by any type of linkages.

It's just a deficiency we live with.
 
Last edited:
  • Like
Reactions: bhzmark
It seems highly unlikely Tesla will charge a steep $8000 at purchase or even steeper $10,000 post-purchase to flip the software switch for the "affordable" Model 3, right? Therefore it makes sense that you should wait until at least July to see if the exact same software will also be cheaper for Model S and X. We have seen all of the angry posts from people who believe they were mistreated and misled by missing the battery-related price cut a couple days ago, so be smart and wait if you are getting ready to add AP2 or FSD.
Sure, if your time is valueless to you and you want to spend several months driving manually - says I as I type this while AP2 pilots me down the road at the end of a 500 mile day that left me relaxed and feeling like I haven't driven even five miles.
 
  • Like
Reactions: bhzmark