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.
More CRAZY ass Elon Tweeting.o_Oo_O Putin is not one to play around with since he plays SUPER dirty and DEADLY. He as had numerous enemies assassinated for trivial actions he deemed inappropriate all over the world. Of course Putin is rather preoccupied and stumped with his inept, poorly trined/motivated and ill equipped military right now but......

🇺🇦🇺🇦🇺🇦
All Ukraine related tweets are better posted here ….

 
I am not sure exactly where the confusion lies

FSD product value proposition presented by Tesla varies vastly depending on when you bought. Even when the name stays the same and Elon's communications fairly consistently refers to original L5 vision.

This confusion leads to endless debates.

Screenshot of the FSD page I shared in this thread was from 2017.
 
  • Like
Reactions: Mrcarcrazy
With regards to FSD it's hard to get more ambitious than the second master plan. Will part 3 be a walk back or will he double down?
From master plan part deux:
Autonomy
As the technology matures, all Tesla vehicles will have the hardware necessary to be fully self-driving with fail-operational capability, meaning that any given system in the car could break and your car will still drive itself safely. It is important to emphasize that refinement and validation of the software will take much longer than putting in place the cameras, radar, sonar and computing hardware.

Even once the software is highly refined and far better than the average human driver, there will still be a significant time gap, varying widely by jurisdiction, before true self-driving is approved by regulators. We expect that worldwide regulatory approval will require something on the order of 6 billion miles (10 billion km). Current fleet learning is happening at just over 3 million miles (5 million km) per day.

I should add a note here to explain why Tesla is deploying partial autonomy now, rather than waiting until some point in the future. The most important reason is that, when used correctly, it is already significantly safer than a person driving by themselves and it would therefore be morally reprehensible to delay release simply for fear of bad press or some mercantile calculation of legal liability.

According to the recently released 2015 NHTSA report, automotive fatalities increased by 8% to one death every 89 million miles. Autopilot miles will soon exceed twice that number and the system gets better every day. It would no more make sense to disable Tesla's Autopilot, as some have called for, than it would to disable autopilot in aircraft, after which our system is named.

It is also important to explain why we refer to Autopilot as "beta". This is not beta software in any normal sense of the word. Every release goes through extensive internal validation before it reaches any customers. It is called beta in order to decrease complacency and indicate that it will continue to improve (Autopilot is always off by default). Once we get to the point where Autopilot is approximately 10 times safer than the US vehicle average, the beta label will be removed.

Sharing
When true self-driving is approved by regulators, it will mean that you will be able to summon your Tesla from pretty much anywhere. Once it picks you up, you will be able to sleep, read or do anything else enroute to your destination.

You will also be able to add your car to the Tesla shared fleet just by tapping a button on the Tesla phone app and have it generate income for you while you're at work or on vacation, significantly offsetting and at times potentially exceeding the monthly loan or lease cost. This dramatically lowers the true cost of ownership to the point where almost anyone could own a Tesla. Since most cars are only in use by their owner for 5% to 10% of the day, the fundamental economic utility of a true self-driving car is likely to be several times that of a car which is not.

In cities where demand exceeds the supply of customer-owned cars, Tesla will operate its own fleet, ensuring you can always hail a ride from us no matter where you are.
 
With regards to FSD it's hard to get more ambitious than the second master plan. Will part 3 be a walk back or will he double down?
From master plan part deux:
Master Plan #3:

1. Implement Master Plan #2:oops:🤣🤣

Screen Shot 2022-03-17 at 6.32.46 AM.png
 
With regards to FSD it's hard to get more ambitious than the second master plan. Will part 3 be a walk back or will he double down?
From master plan part deux:

As the technology matures, all Tesla vehicles will have the hardware necessary to be fully self-driving with fail-operational capability, meaning that any given system in the car could break and your car will still drive itself safely.

Does Elon really think that our cars do this? I guess he does since the website has said our cars have the necessary hardware for FSD for years now. But, considering the times when the cameras sometimes get blinded or blocked, or the computer crashes, I don't see how we can say that our cars have the necessary redundancy now that they could still fully self-drive even if any system failed. Clearly, they don't.
 
Does Elon really think that our cars do this? I guess he does since the website has said our cars have the necessary hardware for FSD for years now. But, considering the times when the cameras sometimes get blinded or blocked, or the computer crashes, I don't see how we can say that our cars have the necessary redundancy now that they could still fully self-drive even if any system failed. Clearly, they don't.


FWIW the computer was specifically designed with the necessary redundancy-- two entirely separate nodes with independent power delivery and such.

But then it turned out HW3 didn't have nearly enough compute power in a single node to run even the current L2 code, so they had to use both nodes for a single instance of the whole stack and thus gave up the redundancy.

I'd expect HW4, at least initially, to get back that redundant capability. Then it's just a question of it THAT is enough compute to do the job in a single node or not.

Since nobody has actually done this before nobody actually KNOWS how much compute is enough after all, so everyone is guessing until it works.


Most of the other systems likewise have some redundancy- somewhat famously Tesla recently REMOVED a redundant steering part in china due to parts shortages (and the fact the cars aren't autonomous right now so it doesn't need the redundancy right now and they could always retrofit later if they had to)


Cameras obviously lack some degree of redundancy, but it's likely one could still manage some type of minimum risk fallback losing a single camera (and AFAIK each camera is wired independently, so one dying shouldn't take out others).
 
Interesting, last year Tesla had 5,760 Nvidia A100s in their cluster. At $10k a piece that's only $58 million. Of course there's more to the cost than just the GPUs but I wonder if they've expanded it?

Anyway FSD is extremely hard but it appears that BMW is planning on supervised self-driving and a highway L3 system. That's not nearly as ambitious as FSD.
 
In fact all the cars have the ultimate redundancy system...us

I thought you were European, but that sounds like a very American thing to say. Unless its a joke and then its funny.

European regulators aren't onboard with the whole advanced L2 thing with having humans as redundancy for the machine. Instead you get features that are trimmed down. Like AP lite instead of AP, and even Auto Lance Change is under review.

Tesla can't even sell FSD there as the Europeans don't like the way Tesla advertises it.

In Europe UltraCruise from GM will turn into Ultranot.
 
  • Like
Reactions: Sporty
Interesting, last year Tesla had 5,760 Nvidia A100s in their cluster. At $10k a piece that's only $58 million. Of course there's more to the cost than just the GPUs but I wonder if they've expanded it?

That is in ONE cluster. Of the 3 mentioned at the time. So you're at $174 million just for GPUs. There's also gonna be a bunch more $ on the networking/connectivity, cooling, enclosures, power supply, etc....

But still probably not gonna get to "multibillion" even then


Anyway FSD is extremely hard but it appears that BMW is planning on supervised self-driving and a highway L3 system. That's not nearly as ambitious as FSD.

To be fair, German car makers have been "planning" L3 systems for many years now and repeatedly failing to actually deliver them.

Audi said it'd be selling L3 in 2017. 5 years later and they're selling no such thing.

Mercedes claimed last year they'd have an L3 system on sale "early 2022".... AFAIK they still claim that but it's not available yet... but will allow L3 driving.... as long as you're on about 8000 miles of specific German highways and traveling 37 mph or less.
 
Phone ?

They have no idea how hard FSD is. Only path to success imo is hardcore real-world AI software with dedicated NN inference acceleration ASICs in car, multibillion dollar NN training supercluster and 10+ billion miles of vehicle data. Good luck.​
Thanks! My phone isn't allowed in my workplace (prison) and it's a 15+ minute walk to my car
 
As the technology matures, all Tesla vehicles will have the hardware necessary to be fully self-driving with fail-operational capability, meaning that any given system in the car could break and your car will still drive itself safely.

Does Elon really think that our cars do this? I guess he does since the website has said our cars have the necessary hardware for FSD for years now. But, considering the times when the cameras sometimes get blinded or blocked, or the computer crashes, I don't see how we can say that our cars have the necessary redundancy now that they could still fully self-drive even if any system failed. Clearly, they don't.
Elon and Tesla love those aspirational statements. "As the technology matures, all Tesla vehicles will have..."

Yes I bet he does believe what he says, those statements certainly could be proven true one day. However what they seem to lack is the conviction to actually do the work necessary to achieve the statement.

Regarding redundancy they need to actually have sensor, computer, and mechanical redundancy to achieve redundancy. Removing steering redundancy for example does not show a global commitment to the goal.

Aspirational statements get picked up by the press and become part of Tesla-lore: “You don’t even need to touch the shifter in new Model S. Auto detect direction will come as an optional setting to all cars with FSD.”. Is that true on new Model S cars, you never have to shift? No it's not true, but widely reported.