-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
10.9 FSD
- Thread starter Buckminster
- Start date
sperkin
Active Member
I just used FSD beta from my house to Big Bear Mountain Ski Resort and back. I was very surprised it handle it well. It did 2 major mistakes, 2 minors, and 2 slow decisions. Overall, I was very impressed.
2 Major mistakes were caused by pickup trucks not following traffic laws and drove like idiots. One was in the left turning lane to skip traffic and cut me off as he bullied his way into my lane which was going straight. As a human, I saw that one coming, but FSD reacted late and swerved to the right and slammed on the brakes. Another major ones, I was going straight and a pickup truck didn't stop at a stop sign and forced his way into my lane to make a left. As a human, I would have already slowed down, but FSD swerved to on-coming traffic, and I took over right before the truck hit me.
2 Minor ones where the car crossed double yellow doing a couple of tight turns where we were all going slower than the speed limit.
2 slow decisions were passing big rigs doing 20 mph, and I was leapped frogged by 20+ cars. I had to manually take over to pass the trucks.
2 Major mistakes were caused by pickup trucks not following traffic laws and drove like idiots. One was in the left turning lane to skip traffic and cut me off as he bullied his way into my lane which was going straight. As a human, I saw that one coming, but FSD reacted late and swerved to the right and slammed on the brakes. Another major ones, I was going straight and a pickup truck didn't stop at a stop sign and forced his way into my lane to make a left. As a human, I would have already slowed down, but FSD swerved to on-coming traffic, and I took over right before the truck hit me.
2 Minor ones where the car crossed double yellow doing a couple of tight turns where we were all going slower than the speed limit.
2 slow decisions were passing big rigs doing 20 mph, and I was leapped frogged by 20+ cars. I had to manually take over to pass the trucks.
Last edited:
CaptUAL
Member
How much did you get paid for risking your car, neck and family? So much for a relaxing week end drive to the mountains. But you have the honor of knowing you have served the Elon Land. I know, it's for the greater good.
Had the update for awhile now and just completed a round trip 4 hr drive. A few things I noticed:
1. I had a number of nag alerts that could not be dismissed by steering wheel tug, thumbwheels or left stalk down (model y). I disengaged and re-engaged fsd each time. This was something that had occurred on earlier builds but seemed to be fixed during later updates. Seems it is back for me.
2. It now appears there is a grace amount of speed > 80 mph. Previously fsd would disengage as soon as I went over 80 mph. This time it let me get a few mph over while I was passing someone. Going way over did get the pink wheel disengage and fsd lockout.
3. Interestingly I switched profiles to a non fsd beta enabled one cheafter the lockout and was able to use regular autosteer/cruise control. Then switched back to the fsd profile and fsd worked. Upon completion of my trip, it did not appear the lockout was counted against me.
1. I had a number of nag alerts that could not be dismissed by steering wheel tug, thumbwheels or left stalk down (model y). I disengaged and re-engaged fsd each time. This was something that had occurred on earlier builds but seemed to be fixed during later updates. Seems it is back for me.
2. It now appears there is a grace amount of speed > 80 mph. Previously fsd would disengage as soon as I went over 80 mph. This time it let me get a few mph over while I was passing someone. Going way over did get the pink wheel disengage and fsd lockout.
3. Interestingly I switched profiles to a non fsd beta enabled one cheafter the lockout and was able to use regular autosteer/cruise control. Then switched back to the fsd profile and fsd worked. Upon completion of my trip, it did not appear the lockout was counted against me.
Absolutely, and sending emails for most of the non-obvious ones, although I'm never sure exactly which of my many snapshots make it to the end of the day to upload over wifiAre you reporting these things using the FSD report button? That's how to get them fixed!
What is your source for this info?
Just for reference Tesla has *70,000 employees world wide. So lets (optimistically) assume about ½ (35K) are in the US. Also clips come in 7 days a week and employees work on average a little under 5 days a week (holidays, vacations, time off, etc). So you are saying that about ¼ of Tesla's entire US staff (Mechanics, Assembly line, Delivery personal, etc.) spends about ½ of their working time watching Beta clips?
That is almost 11,000,000 man-hours a year or about $500,000.00 in cost.
*Tesla SEC filing: “As of December 31, 2020, our full-time count for our and our subsidiaries’ employees worldwide was 70,757..."
Clearly this is not the case.
Last edited:
EVNow
Well-Known Member
I misremembered - it is actually 1,000 labelers.
I think that makes it 300 clips a day per labeler. AT 5 minutes, clearly not possible. They need some way to reduce automatically label those and for labelers to look at only critical ones.
EVJAutoJunkie
EV Junkie
I'm still waiting for 10.9 to drop into my download queue, but the rollout is still at a dribble. There is still no feedback from Tesla as to why it has been limited to a smaller number of users.
Last edited:
More than 2 weeks after it began to roll out, it stopped (at ~40% of testers) and since he/they've been uncharacteristically silent.
Well, either something is broken that they are focusing on, or Tesla is taking a 2 week team building exercise!
The Auto-labeler - which they use. Dojo will allow for higher efficiencies of the auto-labeler. Some labeling is done by humans and some are handled by an algorithm.
I’m wondering if they use the auto labeler to label all videos and the 1k humans are used only to review auto-labeled clips and verify that they‘ve been labeled correctly as a kind of quality control?
I have been monitoring my car’s network traffic, and it doesn’t appear that the camera-clip-button-presses have been uploading data for about a week. This may be just a bug with my car (or maybe they are tired of reviewing my button presses). Anyone seeing the normal 2gig upload with each camera button press, or has it stopped for others?
sperkin
Active Member
There wasn't any risk since my hands were on the wheel. FSD will go over double yellow by a couple of inches and was easy to reel it back in. The pickup truck in front of me by a few cars slammed into the side of the mountain when we were going down. She was looking at the sun set. All air bags deployed.
[conjecture]I bet they only check a few at random, like 1% just for quality control. Probably the Auto Labeling checks for common events like Right Turn on Red at a No Turn on Red intersection (bet they get 1000's of this event each day and easily labeled). It probably flags events that are unusual/edge, violent or that it can't identify and theses are sent to the humans to look over. Likely the identified clips (maybe <>95%) are sorted and aggregated into a "bucket" of similar events (like the Right at a No Turn) then analyzed as a group for improvements or changes that can be made.
I love how there is zero evidence of Tesla ever reviewing any clip or email, but people keep diligently working.
I say this as not being all that different.
I hit the button
I email
In a way its like trying to attract a mate.
You put effort into how you look
You get new clothes
You get a new haircut
You figure if you put in the work that they'll notice you.
You get together with your buddies to work on strategy, and to compare notes. You hope that maybe someone had success.
At the end of the day they never do, and you're left wondering if all that was worth it. Then you wonder if maybe the entire point was just to go through it because it made you feel like you were a part of something despite the fact that you were simply an imposter.
I say this as not being all that different.
I hit the button
I email
In a way its like trying to attract a mate.
You put effort into how you look
You get new clothes
You get a new haircut
You figure if you put in the work that they'll notice you.
You get together with your buddies to work on strategy, and to compare notes. You hope that maybe someone had success.
At the end of the day they never do, and you're left wondering if all that was worth it. Then you wonder if maybe the entire point was just to go through it because it made you feel like you were a part of something despite the fact that you were simply an imposter.
Last edited:
Or more likely the next release will be V11 with the FSD single stack. Total silence from Elon on anything 10.x.
converting "disengagements per mile" to "miles per disengagement" is simply the reciprocal fraction..... You seemed to have intuitively figured it out.
0.22 disengagements per [1] mile = 0.22/1. Reciprocal would be 1/0.22 = ~4.5mi/disengagement. if you just write the fraction the same way as you write the units: disengagements/mile vs miles/disengagements. in the former, disengagements is on top; in the latter, miles is on top.
tmoz
S85D, Y LR, ex Prius PiP
10.0 blew his shoes off. 10.8 blew his socks off. Perhaps he is worried about more clothing items getting blown off by the next release.
Similar threads
