-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
OFFICIAL BUTTON WATCH
- Thread starter Drrobmiller
- Start date
EVNow
Well-Known Member
GargantulaKon
FSD Beta Chill Member
I had that problem a few times. I swiped the app away in the app overview and tried again and then it usually works.
It was on US-1 near me which I believe is as close to a divided highway as you can get without going on an interstate system.
Before today I would turn AP on and off regularly to avoid unwanted system behavior like braking when not necessary for turning cars or slowing at a light (my AP likes to brake hard at the last second for lights, I am much more conservative when slowing).
I wonder if Tesla has pulled the button, given that Musk seems to think they have more than enough people requesting early access already......
So I called Tesla service and they went thru each requirement needed for FSD Beta button, like if there are multi cars in same house, FSD Purchase date,
Computer, build date, power cycling, etc. and they have now escalated this one as they haven’t had any other cars with the same issue. You might be right and they haven’t been communicated yet but
that’s not the case and they get this figured out.
Atari2600
Active Member
If it is based on reality sure. When I drive down the road the correct direction on my street cars often set off the collision alarm. I got my score up from a 3 to an 86 the first day. The next three days I got a 100 each day. Basically I have to drive out of my way when I go anywhere. My overall score is now a 95 as I rack up more miles.
For a statistical model to work it must use valid data!
FalconFour
Member
O dam. OK, something is very wrong.
- Version is 2021.32.22 (same as mine)
- Has NoAP so FSD is enabled
- FSD explicitly says it's enabled
- Release notes don't mention The Button but do mention the changes to Immersive Sound
I'd suspect somehow your car might be configured for an alternate build of some kind... like how Canada doesn't get it, for example. You're not in Canada are ya? lol
Last edited:
boonedocks
MS LR Blk/Blk 19”
Multi cars don’t matter and neither does year / buy / subscribe either. Something else completely I M sure!So I called Tesla service and they went thru each requirement needed for FSD Beta button, like if there are multi cars in same house, FSD Purchase date,
Computer, build date, power cycling, etc. and they have now escalated this one as they haven’t had any other cars with the same issue. You might be right and they haven’t been communicated yet but
Silicon Desert
Well-Known Member
Thanks. I was able to have the score rounded up today. How do you think the overall mileage is weighted? The app only shows the day's total miles but not for individual trips' miles on the day. It seems like other than FCW, the denominator for the individual factors only includes "applicable miles," e.g., total miles across all days where there's braking >0.1g.
I forget the numbers I checked, but something like:
Day 1: 2% hard braking, 10 miles
Day 2: 0% hard braking, 10 miles
Overall hard braking was not 1%. It also seems like if Day 2 was mostly on Autopilot, those Autopilot-braking-miles are not counted (nor are they counted in the numerator if it was hard) resulting in an overall hard braking closer to 2%.
I would guess the day's score does the same thing for aggregating trips.
When using autopilot correctly, you won’t get dinged if the car brakes too hard or follows another car too close. It won’t count any of the braking as braking time, or any of the following as following time (unless you press the accelerator I believe). The miles do count though when it averages your scores across each day to give you your average score.
BitJam
Active Member
A statistical model will be right on average but not right in every case. There are probably some cases where it's way off. The model not working in your case does not imply the 5 billion miles of fleet data they used to build the model are invalid.
People seem to be under the false impression that the safety test is supposed to be fair. It's not about being fair, it's about weeding out bad apples even if that means weeding out some good drivers. As long as Tesla gets enough beta testers, weeding out unlucky good drivers is not an issue for them.
Neural nets like the ones Tesla uses for FSD need a massive amount of input data for training. This is why Tesla is building the world's largest AI computer. It's also why they hired a team of 1,000 people to handle things computers have trouble with. OTOH the NTSB has already blamed Tesla for accidents caused by inattentive drivers and said in no uncertain terms that Tesla should not give the beta to more drivers until the safety issues have been worked out.
Tesla needs a large pool of beta testers to get the data they need to improve FSD. But they can't let the beta process appear unsafe. I thought their safety test was a brilliant way out of this Catch-22. It allows them to safely (I hope) expand the pool of beta testers so they can rapidly improve FSD. If the accident rate is not vastly lower than the overall average then the government will likely shut down Tesla's FSD program which would be disastrous on many levels. Ideally, as Tesla expands the pool of beta testers FSD will get better and safer thus making it safe for them to further expand the pool. This is mind blowing!
For the safety test it appears Tesla did a large linear regression (I imagine they inverted a big matrix using SVD) to find out which measurables correlated most strongly with collisions. Then to make things easier on humans they reduced it to just the five variables that had the highest correlations. This seems like an excellent approach, especially for their first beta. I can't think of a better way. But perhaps you can. If so, please chime up because Elon is often open to good suggestions.
I guess I should probably avoid valet parking for the next two weeks.
I wonder if valet mode with no connected phone key counts towards the score...
Similar threads
