-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
It'll be a one-size-fits all.
Also, what's the car going to do if there's a police car approaching you to pull over? AP1.0 doesn't use the rear camera. Unless they changed this on AP2.0, it wont be FSD capable. Also, there are camera dead zones (even from the picture on the website).
Or do you mean in general when other car manufactures get to Level 3/4? I think one size fits all also.
Also, what's the car going to do if there's a police car approaching you to pull over? AP1.0 doesn't use the rear camera. Unless they changed this on AP2.0, it wont be FSD capable. Also, there are camera dead zones (even from the picture on the website).
Or do you mean in general when other car manufactures get to Level 3/4? I think one size fits all also.
You're saying you think that the Tesla engineering team possibly forgot that a FSD car needs vision to the rear and might have neglected to use the rear facing camera in the new hardware suite?
You're right, the Tesla engineers have never over promised and under delivered.
Though I doubt it'd be something this simple.
I'm willing to bet on just about anything feasible that AP2.0 will not be fully self driving. Level 3? Sure. Level 4 or 5? No way.
It will not be Level 5 autonomous as Elon promised. Have you seen the description of Level 5?
It will not be Level 5 autonomous as Elon promised. Have you seen the description of Level 5?
- Level 5: Other than setting the destination and starting the system, no human intervention is required. The automatic system can drive to any location where it is legal to drive.
- Level 4: The automated system can control the vehicle in all but a few environments such as severe weather. The driver must enable the automated system only when it is safe to do so. When enabled, driver attention is not required.
Sorry, one more piece of info, then I'll step down from the soapbox.
NHTSA abandoned it's 4 levels of autonomous driving and adopted SAE's 5 levels a couple months ago. Link
So, add to that this one key piece of information for vehicle performance guidelines "The process should describe design redundancies and safety strategies for handling cases of HAV [Highly Automated Vehicle] system malfunctions." Link from NHTSA Autonomous car guide [emphasis mine]
HAV seems to apply to Level 3 too...
NHTSA abandoned it's 4 levels of autonomous driving and adopted SAE's 5 levels a couple months ago. Link
So, add to that this one key piece of information for vehicle performance guidelines "The process should describe design redundancies and safety strategies for handling cases of HAV [Highly Automated Vehicle] system malfunctions." Link from NHTSA Autonomous car guide [emphasis mine]
HAV seems to apply to Level 3 too...
Dude the soapbox is old and the 3/4/5 horse has been beaten to death. I don't think anyone here on the forum claims/believes that this car will drive in all weather conditions nor even that it has the redundancy regulators might require to allow drivers to legally shut their eyes.
What I do think would be completely absurd is if the car has no vision to the rear just like the 1.0 suite - that's beyond underdelivering and isn't reasonable at all by any definition of self driving. The car has to be able to check behind it for approaching traffic or it can't change lanes. That's the limitation of 1.0 we all know. You are theorizing that Tesla shoved in 8 cameras, a 12 trillion operations per second computer to run the neural network. Then Tesla trumpets it to the entire world, claims the cars will drive themselves in a ride share network, and starts publishing videos of the cars driving themselves in city traffic.
But then Jim over in cubicle three forgets to double check his specs and they don't wire up vision to look at traffic behind the car.
That theory is patently absurd. AP 1.0 was bleeding edge at the time (still is in terms of lane keeping performance). This new system is more than an order of magnitude more powerful and you think it still might not even look behind it?
What I do think would be completely absurd is if the car has no vision to the rear just like the 1.0 suite - that's beyond underdelivering and isn't reasonable at all by any definition of self driving. The car has to be able to check behind it for approaching traffic or it can't change lanes. That's the limitation of 1.0 we all know. You are theorizing that Tesla shoved in 8 cameras, a 12 trillion operations per second computer to run the neural network. Then Tesla trumpets it to the entire world, claims the cars will drive themselves in a ride share network, and starts publishing videos of the cars driving themselves in city traffic.
But then Jim over in cubicle three forgets to double check his specs and they don't wire up vision to look at traffic behind the car.
That theory is patently absurd. AP 1.0 was bleeding edge at the time (still is in terms of lane keeping performance). This new system is more than an order of magnitude more powerful and you think it still might not even look behind it?
Round and round we go?
And looking at the diagram, it's obviously not the case. They said they went from 1 to 4 cameras (1 = forward camera, 4 = 3 forward, 1 rear). The 8 cameras include the 4 on the sides.
CHG-ON
Still in love after all these miles
Oh boy! I pray for options. I WANT to drive my cars. If it really does go to becoming an appliance like a refrigerator with tires (I had Volvo that looked like one!), I will be one mighty unhappy camper. My dream is that the car would be completely autonomous or somewhere in between, where it could intervene when needed, depending on user choice.
Honestly, I think the forecasts of this happening in the next few years are way too optimistic. There are so many complexities in this topic. Such as the car needs to hear in addition to see, from every angle, and discern what it is seeing and hearing. Think ambulance...And let's not forget about extreme weather conditions. Think about the human reasoning it takes in order to navigate a blizzard when trying to get home to your loved ones. No, we probably should not drive through blizzards and hurricanes. But that is what humans do. That is what I have done and I know that I am not alone.
I simply do not see that we have the computing power, the programming expertise or the real world modeling to have this become a reality for some time. How much time? I do not know. But I will bet dollars to donuts, it is not within the next 5-10 years.
I fear that we humans are trusting machines too much. They can be useful tools under our command and control. When we allow them to take control, which I fear we are willing to do, we are, plain and simple, F--ked.
Read Elon Musk and Stephen Hawking on AI. And read it carefully.
Honestly, I think the forecasts of this happening in the next few years are way too optimistic. There are so many complexities in this topic. Such as the car needs to hear in addition to see, from every angle, and discern what it is seeing and hearing. Think ambulance...And let's not forget about extreme weather conditions. Think about the human reasoning it takes in order to navigate a blizzard when trying to get home to your loved ones. No, we probably should not drive through blizzards and hurricanes. But that is what humans do. That is what I have done and I know that I am not alone.
I simply do not see that we have the computing power, the programming expertise or the real world modeling to have this become a reality for some time. How much time? I do not know. But I will bet dollars to donuts, it is not within the next 5-10 years.
I fear that we humans are trusting machines too much. They can be useful tools under our command and control. When we allow them to take control, which I fear we are willing to do, we are, plain and simple, F--ked.
Read Elon Musk and Stephen Hawking on AI. And read it carefully.
Similar threads
