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

FSD Beta 10.69

This site may earn commission on affiliate links.
Why are they even bothering with this release? I don’t get it.

Chuck Cook's speculation was that the firmware updates are additive, not cumulative, so in order to prepare us for the holiday release on 2022.44, we first need to receive the FSD Beta branch of 2022.40.

10.69.3.3 might not even contain any new FSD code, it may literally just be a merging and validation of the last FSD code into the 2022.40 branch.
 
Why are they even bothering with this release? I don’t get it.
That is the big question especially if released to only some in the fleet.

There aren't any new features and it wastes team resources to fix bugs on a possibly dead-end software path. The worst case scenario is if the 69 path isn't a dead-end and instead is being fed into the V11 path. If so it's more of the same performance until the FSD team wakes up with new blood/innovation/leadership/enema.
 
  • Like
Reactions: AlanSubie4Life
  • Informative
Reactions: Phlier
Chuck Cook's speculation was that the firmware updates are additive, not cumulative, so in order to prepare us for the holiday release on 2022.44, we first need to receive the FSD Beta branch of 2022.40.

10.69.3.3 might not even contain any new FSD code, it may literally just be a merging and validation of the last FSD code into the 2022.40 branch.
This theory is very possible. My Model Y was in the body shop for about 2 weeks. When I got it back 2 days ago, I did the usual test everything, check everything, drive around, do a two scroll wheel restart. After the restart an update came up to 2022.40.4.2 (from 2022.40.4.1) so I installed it. About 5-10 mins later after the install was complete, I got notified that I had an update. For a second I thought it was a glitch, but after checking the app it prompted me to update to 2022.44.2.

Now my non-FSD Model Y is on 2022.44.2 and my FSD “more premium” Model S is on 2022.36.20 with -$10k in my pocket 😅😅
 
This theory is very possible. My Model Y was in the body shop for about 2 weeks. When I got it back 2 days ago, I did the usual test everything, check everything, drive around, do a two scroll wheel restart. After the restart an update came up to 2022.40.4.2 (from 2022.40.4.1) so I installed it. About 5-10 mins later after the install was complete, I got notified that I had an update. For a second I thought it was a glitch, but after checking the app it prompted me to update to 2022.44.2.

Now my non-FSD Model Y is on 2022.44.2 and my FSD “more premium” Model S is on 2022.36.20 with -$10k in my pocket 😅😅
I’m going to go reboot the S after reading this 😀. The 3 is on 2022.44.2 and I really want/need Zoom in Lucille. Abigail has no need for Zoom and the wife doesn’t give 2 shts about half the stuff coming.

What really kills me is she doesn’t like Rainbow road.
 
  • Funny
Reactions: Ramphex
This theory is very possible. My Model Y was in the body shop for about 2 weeks. When I got it back 2 days ago, I did the usual test everything, check everything, drive around, do a two scroll wheel restart. After the restart an update came up to 2022.40.4.2 (from 2022.40.4.1) so I installed it. About 5-10 mins later after the install was complete, I got notified that I had an update. For a second I thought it was a glitch, but after checking the app it prompted me to update to 2022.44.2.

Now my non-FSD Model Y is on 2022.44.2 and my FSD “more premium” Model S is on 2022.36.20 with -$10k in my pocket 😅😅
I took that as an indication that once your car has "noticed" there is an offered release it must install that before getting the next one.
I've seen that a couple of times on my car where it has a new release on following days
I'll be glad when I get off 10.69.3.1, I'm bored of other drivers thinking I'm drunk.
 
  • Disagree
  • Like
Reactions: KArnold and Ramphex
I took that as an indication that once your car has "noticed" there is an offered release it must install that before getting the next one
There's been quite a few reports of people not installing 2022.44.x even after the vehicle downloaded it and staying on 2022.40.x instead then successfully installing FSD Beta 10.69.3.3 / 2022.40.4.10. This same pattern happened for 10.69.3.1 / 2022.36.20 for people who stayed on 2022.36.x instead of updating to 2022.40.x.

The speculation of additive software doesn't really match up with what has been observed even looking at the previous versions updating to the current FSD Beta version on TeslaFi:
Code:
2022.4.5.21 (10.11.2)     1
2022.20.8                 3
2022.20.9                 4
2022.20.17 (10.69.2.2)    1
2022.24.1                 1
2022.28.300               1
2022.36.6                 2
2022.36.20 (10.69.3.1)  122
2022.40.4                 6
2022.40.4.1              43
2022.40.4.2              71
2022.40.4.5 (10.69.3.2)   2

The majority of updates so far have been from non-FSD-Beta versions across a variety of 2022.20.x, 2022.24.x, 2022.28.x, 2022.36.x, and 2022.40.x. Similarly even for the FSD Beta updates, there's a variety of previous versions that have "skipped" intermediate versions and instead getting a "cumulative" update. Or another words these updates are self-contained and doesn't depend on what was in the previous version. In the rare case of FSD Beta 10.3's rollback for the false automatic emergency braking bug, Tesla pushed out a previous version instead of "subtracting."

Separately, looking at the updates so far, estimating the remaining pending installations to be like those who have already installed, the current rollout has been available to about 5% of the existing FSD Beta population while similarly expanding the population with new testers by about 5%. Previously TeslaFi's FSD Beta fleet was about 4800, and now it's closer to 5000.
 
74 year old here, I don't have a problem with correcting FSD and it is helpful on long drives. I know as I get older I will lose more of my driving abilities so I am participating in this experiment with the hope of some day realizing the advertised potential of full self driving so I can continue to be active and safe.
OTOH, 79 year old here. I agree with above. Eagerly awaiting the 69.3..3 and v11 releases
 
If so it's more of the same performance until the FSD team wakes up with new blood/innovation/leadership/enema.
We'll see mostly small changes that accumulate over time resulting in better FSDb. There may be concentrated effort to get something to a step change (CULT) .... but I'm not in the group that beleives "exponential" changes are somehow around the corner.
 
We'll see mostly small changes that accumulate over time resulting in better FSDb. There may be concentrated effort to get something to a step change (CULT) .... but I'm not in the group that beleives "exponential" changes are somehow around the corner.
They "document" a lot of changes and use percentages to indicate how they are taking measurements.

  • – Upgraded the Object Detection network to photon count video streams and retrained all parameters with the latest autolabeled datasets (with a special emphasis on low visibility scenarios). Improved the architecture for better accuracy and latency, higher recall of far away vehicles, lower velocity error of crossing vehicles by 20%, and improved VRU precision by 20%.
  • – Converted the VRU Velocity network to a two-stage network, which reduced latency and improved crossing pedestrian velocity error by 6%.
  • – Converted the NonVRU Attributes network to a two-stage network, which reduced latency, reduced incorrect lane assignment of crossing vehicles by 45%, and reduced incorrect parked predictions by 15%.
  • – Reformulated the autoregressive Vector Lanes grammar to improve precision of lanes by 9.2%, recall of lanes by 18.7%, and recall of forks by 51.1%. Includes a full network update where all components were re-trained with 3.8x the amount of data.
  • – Added a new “road markings” module to the Vector Lanes neural network which improves lane topology error at intersections by 38.9%.
  • – Upgraded the Occupancy Network to align with road surface instead of ego for improved detection stability and improved recall at hill crest.
  • – Reduced runtime of candidate trajectory generation by approximately 80% and improved smoothness by distilling an expensive trajectory optimization procedure into a lightweight planner neural network.
  • – Improved decision making for short deadline lane changes around gores by richer modeling of the trade-off between going off-route vs trajectory required to drive through the gore region
  • – Reduced false slowdowns for pedestrians near crosswalk by using a better model for the kinematics of the pedestrian
  • – Added control for more precise object geometry as detected by general occupancy network.
  • – Improved control for vehicles cutting out of our desired path by better modeling of their turning / lateral maneuvers thus avoiding unnatural slowdowns
  • – Improved longitudinal control while offsetting around static obstacles by searching over feasible vehicle motion profiles
  • – Improved longitudinal control smoothness for in-lane vehicles during high relative velocity scenarios by also considering relative acceleration in the trajectory optimization
  • – Reduced best case object photon-to-control system latency by 26% through adaptive planner scheduling, restructuring of trajectory selection, and parallelizing perception compute. This allows us to make quicker decisions and improves reaction time.
  • – Introduced foundational support for model-parallel neural network inference by sharing intermediate tensors across SOCs to improve road edge and road line prediction consistency through changes to TRIP compiler, inference runtime, and inter-processor communication layer.
  • – Improved handling of traffic control behavior in dense intersection areas by improving the association logic between traffic lights and intersections
 
They "document" a lot of changes and use percentages to indicate how they are taking measurements.
Right ... all this results in gradual improvement. Nothing exponential ... that would suddenly take disengagement rate from 1 in 10 miles to 1 in 100 miles.

I won't be surprised if the disengagement rate crawls to about 1 in 100 miles in 3 to 5 years and tops out. That will be a great ADAS tool but not a robotaxi.
 
  • Like
Reactions: Phlier
We'll see mostly small changes that accumulate over time resulting in better FSDb. There may be concentrated effort to get something to a step change (CULT) .... but I'm not in the group that beleives "exponential" changes are somehow around the corner.
I am hopeful for a step change on the highway though with V11 once the bugs are released with a few point releases.