pjw65
Member
You could just make a Hotspot with your Mobile Phone while Parked somewhere (15-30 Minutes is enough for an ordanary update, not for the Maps!).
This helps a lot! Or park near a public WiFi (MacD for ex.).
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
You could just make a Hotspot with your Mobile Phone while Parked somewhere (15-30 Minutes is enough for an ordanary update, not for the Maps!).
This helps a lot! Or park near a public WiFi (MacD for ex.).
What happened to the 1+ billion fleet miles they've been collecting over the past two years? You know, that giant treasure trove of data that gives them such a competitive advantage over everybody else? They can just go back and dig into all that data and replay what happened, except with the newest AP software, and see how what AP does differs from what the human driver did, right? Should be easy.
Except... they don't actually have all that data, they only have a few snippets here and there.
Yes yes yes of course. But that's not the same as "autopilot thinks this is an opportunity to change lanes with the CURRENT neural network, what does human think" so yeah no point bringing up that old ship.
It's the narrative that @strangecosmos has been shoveling down our throats for the past 2 years, constantly at every moment. Yet there is absolutely no proof of any such thing and all the evidence points to the exact opposite.
Its baffling!
Maybe not, but still much better than all the other toddlers in his neighborhood still struggling to walk.
You dont need 10 million miles to see it fail. You only need 10 (and that's being generous). Let's be honest they released a pre-alpha software
Ah, so you actually have a Tesla now? Or are you just pretending to know how long it typically goes between failures?
You;re saying this on what evidence? Someone's anecdote. Why don't we wait for more videos to start rolling in before we draw any conclusions on how good the software is
Pretty sure his entire argument strategy has been based on pretending to know where Tesla and Mobileye are in their development cycles
From what I’ve seen, it’s largely taking Mobileye’s word for it on every single possible claim they make, and comparing that exclusively against what Tesla has already released to non-EAP customers.
In addition the NEO-M8L does not have integrated RTK like their newer models this would give it a 1.5CM precision. (Including a long list of caveats at the bottom of the datasheet)
Tesla can implement RTK on their end since they have the required sensors mostly accelerometer and gyroscope/IMU.
Wait you are telling me his not part of Mobileye/Intel Marketing/PR department?
It's like he has mobileye Newswire/press releases on text notification straight to this forum.
Lol... it’s public knowledge what the numbering scheme was. If it’s changed, no clue. But, yeah, all this EA posting is really too bad. And as I say, you should know dang well if you are or aren’t in the program.
Doesn’t really work that way TBH. Acknowledging a lane change suggestion doesn’t acknowledge it as an optimal choice as is implied by the feature, nor does it acknowledge that it is safe. You can acknowledge a lane change with someone in your blind spot and the car will not execute the change until it is safe. So in the end having driver intervention will add nothing to the testing unless a complete failure occurs and you need to shut off AP due to a massive failure or dangerous situation.
Disagreed. The car would not even ask you to change lane if there is a car in your blind spot. Thus, your example is not a real life example.
Machine training is based on learning human behaviors under similar situations. Compare against the suggested behavior by the car is a critical step. Driver intervention is telling the car you don't like its suggestion. Repeat over and over again until a pattern is established. When your car prompts you to change lane, it means it believes it is safe. If you don't confirm, it should analyze the surrounding and try to understand why you don't agree...maybe it suddenly becomes unsafe, maybe too late, maybe the driver wants to go somewhere, etc. Or if you confirm and then you suddenly intervene because someone might have got into your lane at the last second, etc. So many possibilities. But it is a "teaching moment" for the car.
At the end, machine learning is not based on what you do. It is based on what vast majority of people would do under a specific scenario. This is exactly why Elon said it needed 10 mil driven miles before he can remove the lane confirmation restriction. Whether 10 mil driven miles is enough or not is up to the experts to decide. However, it is an unavoidable step. You may not like what it does at this time. New owners next year may love it because most of the problems may be resolved by then because we, the early owners, are part of the beta testing group. Remember, it is called Machine Learning....it takes time to "learn"
For those who don't want to be part of the journey, they really should not get EAP. We can go back and debate whether Tesla should have sold a unfinished feature, but it is a dead topic. It is what it is. We learned our lesson. At this point, we should wait for the feature to mature.
I don't think the purpose of the confirmation is for lane change purposes.
It makes more sense for it to be for route purposes.
Few questions they should be able to answer based on the :
- Did I suggest this lane change for taking an exit/following a route too early?
- Did I pick the right of the two exit lanes?
- Was the fastest lane suggestion really faster?
- Is it a faster lane but there is another reason why the driver didn't accept it? (Bunch of trucks ahead on the faster lane)
It has varied from .9 miles to 2.5 miles
OK first of all I have had Nav on AP for about a month and it does suggest a lane chaneg with cars in your blind spot, care to tell me what your experience is? What? Crickets? Thought so....
Second of all machine learning has nothing to do with comparing human, imperfect choices against neural net decisions. It's about comparing outcomes based on decisions to provide better decisions. As I stated in both examples the human intervention provides no data whatsoever about the quality of the outcome and therefore cannot be used in a machine learning approach.
Feel free to post some more baseless nonsense though, I find it pretty funny myself.
Well you still do have the luxury of not enabling any of these beta features, so to me that's the best option for us in the face of unfinished features - though someone will argue it's dangerous, some clown will video themselves driving from the back seat and so on, but you can't please everyone all the time.
Similar threads
