Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
The next big milestone for FSD is 11. It is a significant upgrade and fundamental changes to several parts of the FSD stack including totally new way to train the perception NN.

From AI day and Lex Fridman interview we have a good sense of what might be included.

- Object permanence both temporal and spatial
- Moving from “bag of points” to objects in NN
- Creating a 3D vector representation of the environment all in NN
- Planner optimization using NN / Monte Carlo Tree Search (MCTS)
- Change from processed images to “photon count” / raw image
- Change from single image perception to surround video
- Merging of city, highway and parking lot stacks a.k.a. Single Stack

Lex Fridman Interview of Elon. Starting with FSD related topics.


Here is a detailed explanation of Beta 11 in "layman's language" by James Douma, interview done after Lex Podcast.


Here is the AI Day explanation by in 4 parts.


screenshot-teslamotorsclub.com-2022.01.26-21_30_17.png


Here is a useful blog post asking a few questions to Tesla about AI day. The useful part comes in comparison of Tesla's methods with Waymo and others (detailed papers linked).

 
Last edited:
It won’t matter. It will still be rubbish. I really do think anyone paying for this crap is throwing their money down the toilet. And that’s from someone who bought it three times. Shame on me.
Likely. I’m half smarter than the average bear. Bought it twice, third go around, bought “enhanced AP”, subscribed to FSD for $99/month. Buying enhanced AP was a waste too, car doesn’t even have “dumb” summon, which is part of enhanced AP. Oh well.
 
  • Like
Reactions: jebinc
^I wonder if us on HW4 will get v12 anytime soon, or six months from now?
The Notateslaapp webbsite shows that V12 FSD (2023.38.10) will be delivered to those cars that are on 2023.38.X (FSD 11.4.4. Unfortunately these cars are not HW4. If you ask Elon, he would say that HW4 cars will get it 2 weeks after FSD v12 is delivered to 11.4.4 cars.

You are probably right saying that it may be 6 months before we get V12.
 
The Notateslaapp webbsite shows that V12 FSD (2023.38.10) will be delivered to those cars that are on 2023.38.X (FSD 11.4.4. Unfortunately these cars are not HW4. If you ask Elon, he would say that HW4 cars will get it 2 weeks after FSD v12 is delivered to 11.4.4 cars.

You are probably right saying that it may be 6 months before we get V12.
Is it possible cars with HW3 and SW 2023.27.7/FSD 11.4.7.3 won't get V12 then?
 
  • Funny
Reactions: jebinc
Hilarious yes.
I believe Omar is now stuck with 11.4.4 since he has a new HW4 car. His drives are worse now and you can tell he's annoyed since his older car ran 11.4.7.3 which was better. My new HW4 car is on 11.4.4 and has lots of regressions.
As I posted back in September, some new HW4 cars are running the same version as the 2022 FSD tester group. My son's MY has FSD only because of the free trial period (I think it ends in mid January for him), and he got 2023.7.30 two or three weeks after purchase, which then updated to 2023.27.7 = 11.4.7.3 ((a day or two before I got that myself). I wasn't expecting his car to get that, but it certainly did.

This doesn't seem to apply to all HW4 cars, but it's quite possible that Omar has it also.

I think it's a separate question as to whether these cars will be capable of operating the first release of v12. Originally I thought they might be the first to get it, but evidence and Elon's comments indicate that they might lag the HW3 deployment of v12.
TBD.
 
Hilarious yes.
I believe Omar is now stuck with 11.4.4 since he has a new HW4 car. His drives are worse now and you can tell he's annoyed since his older car ran 11.4.7.3 which was better. My new HW4 car is on 11.4.4 and has lots of regressions.
My HW4 is on 11.4.7.3 and I can confidently say that is has regressed further from 11.4.4 which regressed from 11.4.1
 
  • Like
Reactions: FSDtester#1
Was just stating the facts. OP exaggerated. L2 to L2 features compared was my point. Of course FSDj is trying to do more, but even the basic ones are unpredictable.
FSD is L2. Does Subaru do end to end in city with supervision ? Because that feature is L2.

That is why levels are stupid - they don’t really let you compare consumer cars at present. They are all L2.
 
  • Like
Reactions: FSDtester#1
Yes, like most of your posts Dr.!
Well, I've posted approx 3,800 times in these forums. I'm most honored that you have carefully read them all to generate this essential statistic!

Joking aside, it seems to me that people who post stuff like "it's clear HW3 will never work at L3" or "it will still be rubbish" seem to reflect an attitude that FSDb is "just another" software development effort, like a new release of Windows or iOS. But of course it isnt, it's an attempt to do something that has never been done before using cutting edge technology. No-one knows if the project will succeed, and it's a huge bet by Tesla, and, to a lesser extent, by any and all of us who have paid for FSD. No-one knows how long it will take, or how many blind-alleys it will go down on the way. Sure, it's frustrating, and I'm frustrated too, but channelling that frustration into wild declarations about the technology, when no-one knows what can really be done seems to me as foolish as Elons wildly optimistic pronouncements.
 
Last edited:
FSD is L2. Does Subaru do end to end in city with supervision ? Because that feature is L2.
That is why levels are stupid - they don’t really let you compare consumer cars at present. They are all L2.
Again, I was comparing the reliability/predictably of Subaru’s L2 list to the equivalent Telsa L2 offering. Can’t be any more clear than that.
 
Just finished another round trip up to San Francisco from LA. Route involve 405,5,152,101,85,280,1,35. The most challenging Route is 152 which involves 3 sections: A two lanes each way quite straight divided section, A two lanes each way divided hilly section and A one lane each way undivided local highway section with local intersections. Previously AP, EAP codes have problem with the curvy hilly section and local winding section. The main problems were speed modulations around curves. The current version ran through with no problem at all! The remaining problem is the interchange from 5N to 152W where the exit cut across an entrance. I have been on this route since 2020. The gradual improvements have been satisfying.
 
Well, I've posted approx 3,800 times in these forums. I'm most honored that you have carefully read them all to generate this essential statistic!

Joking aside, it seems to me that people who post stuff like "it's clear HW3 will never work at L3" or "it will still be rubbish" seem to reflect an attitude that FSDb is "just another" software development effort, like a new release of Windows or iOS. But of course it isnt, it's an attempt to do something that has never been done before using cutting edge technology. No-one knows if the project will succeed, and it's a huge bet by Tesla, and, to a lesser extent, by any and all of us who have paid for FSD. No-one knows how long it will take, or how many blind-alleys it will go down on the way. Sure, it's frustrating, and I'm frustrated too, but channelling that frustration into wild declarations about the technology, when no-one knows what can really be done seems to me as foolish as Elons wildly optimistic pronouncements.
A lot of stuff we don't know but the one thing we do is that the car could self-park when it was shipping with USS and cannot do it now (I've posted a couple of times on this, won't do it again now). I get it that self-driving has never been done before, and it will be super impressive if Tesla gets there with just the cameras, but parking sensors, rain sensors, etc - all this has been done before. Why not keep this tried and tested - and cheap - tech until you have a vision-only solution?
 
I apologize for what happened. I went back about 30 minutes later to take the 2nd picture since a car came up behind me. Thought I was at the same spot in the road. Obviously not. I will do this again the right way.
Here are the updated pictures taken at the same time. Hope this helps.

Picture 1- Yellow line shows what the B-pillar can see (yellow line) compared to the driver leaning forward can see. This likely explains why FSD sometimes begins to enter the intersecting requiring an intervention because the driver see's a car coming. FSD simply didn't see the vehicle. Notice how far the van has moved towards the center line since the car is already protruding into the crossing road. FSD never creeps this far by itself so what the B-pillar camera actually sees is worse than these photos show.

What B-Pillar Camera Can See Photo3.jpg


Picture 2- B-pillar camera view. Note same shadows.

B- Pillar Camera View- 2nd Photo.jpg