In the Autonomy Day livestream Elon said the public will demand autonomous cars instead of 2 ton death machines. Well they will take my steering wheel away when they pry my cold dead fingers from it.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
2 ton death machines
- Thread starter electricar
- Start date
diplomat33
Average guy who loves autonomous vehicles
That's funny. You can resist all you want but when AVs become magnitudes safer than human drivers, you WILL lose the right to drive because letting you drive would KILL people.
Daniel in SD
(supervised)
I wouldn't worry too much about it until someone actually makes an autonomous vehicle that's safer than a human. They've been just around the corner for a decade. I expect we'll see fusion power plants first.
I usually like Elon, but when he delivered straight-faced "we will just delete the steering wheel" line I started laughing. This is so far removed from what is probable in the next 10 years - just think of what would be required to actually do that:
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020.
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020
.
Daniel in SD
(supervised)
You forgot the first stepI usually like Elon, but when he delivered straight-faced "we will just delete the steering wheel" line I started laughing. This is so far removed from what is probable in the next 10 years - just think of what would be required to actually do that:
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020
0. Make FSD. Something that no one has managed to do!
Waymo already has a permit to test cars on public roads in California without a steering wheel so I think your timeline for steps 1-11 is wrong.
Permit Holders (Driverless Testing)
You are correct. Probably condense it down to fewer steps, but not to fewer years.
Daniel in SD
(supervised)
Yeah, the number of total years sounds plausible!
I just don't think regulatory approval is as big an obstacle as many around here think.
ATC@LWSK
Member
I usually like Elon, but when he delivered straight-faced "we will just delete the steering wheel" line I started laughing. This is so far removed from what is probable in the next 10 years - just think of what would be required to actually do that:
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020
I'm pretty sure everyone in the space industry laughed 10 years ago when Elon said with a straight face he'll make rockets that will land vertically on the launch pad and be ready for reuse shortly thereafter.
Daniel in SD
(supervised)
Bad example. Not to denigrate SpaceX's achievement but vertical landing rockets have been around for a long time. No one has ever made anything close to a level 5 autonomous vehicle.
But this is not what he said - your quoted 10 years I mean. It sounded like I could plan to take a robotaxi without a steering wheel by 2020 - hence the laugh. I am ready to put my money where my mouth is on this one. Should we wager $100 to be paid to the charity of the winner's choice if by 12/31/2020 there is no robotaxi (even one with a steering wheel
) in any city in US - and I mean not even as a pilot program?As I mentioned above, it took NHTSA 5 years to approve laser matrix lights. So I am not sure we will be in a position to lend our cars to the Tesla network any earlier than 10 years.
Guns still continue to kill tens of thousands of people and we don’t do anything about that so what makes you think we will do anything about cars - we will just have an National Driving Association ( NDA ) and nothing will change.
He also said at the end that he doesn't think that these people that want to ban it will actually succeed nor does he agree with that position. I think he realized that he went a little too far in trying to please his autonomy investor audience and walked it back a little. I think the eventual end game for the human driving vs computer driving relationship is simply the full self driving software still running in the background even when you're driving and when it detects a significant mistake, it will subtly but instantly turn itself on, fix the problem, and then fade back off. This way it allows people to keep driving if they want/need to nut it will make it just as safe as full AV so best of both worlds. They could also have cameras inside the car that track your eyes that automatically turn on AV when you are not paying attention.
I usually like Elon, but when he delivered straight-faced "we will just delete the steering wheel" line I started laughing. This is so far removed from what is probable in the next 10 years - just think of what would be required to actually do that:
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020
Interesting...
1,2,3,4,5 and some degree 6 are already complete, between Tesla and other players in the space. 7,8,9 has been demonstrated fairly well. 10 and 11 are easy after you get 9.
I usually like Elon, but when he delivered straight-faced "we will just delete the steering wheel" line I started laughing. This is so far removed from what is probable in the next 10 years - just think of what would be required to actually do that:
1. Prove that the FSD works - perhaps in a real-life, but geofenced testing environment
2. Get approval to role out a pilot FSD testing program - most likely also in a geofenced area.
3. Provide substantial evidence to the regulators that FSD is superior to human driving in the said test area. I would imagine a 1-month worth of data will not be enough - will probably need at least 1 year if not more.
4. Get approval to expand FSD outside of geofenced area - also on a test basis. This can go in parallel in different states, but I suspect with different adoption timelines.
5. Go back to the regulators with again 1 year or so worth of data to obtain full approval for FSD use = again state by state - with different adoption timelines.
6. Once FSD is fully approved - which up to this moment is still under direct driver control and on the driver's responsibility - start a pilot testing with no driver input.
7-9. Repeat steps 3-5 with FSD but no driver input
10. Convince the legislators that they should approve a car without a steering wheel
11. Find people who are willing to trust their life to the FSD computer and purchase said cars without a steering wheel.
NHTSA is known to act quickly at industry demands - it only took 5 years to approve the matrix adaptive driving high beams. Virtual side-mirrors are also a very quick adoption story. Sure, it's all going to happen in 2020
And the underlying assumption is that this will all actually work well. I'm not at all convinced of that. There are way way too many unique and bizarre situations that crop up in everyday driving that will not be coded or trained for. They require the larger contextual understanding that the human brain brings to the table.
One other interesting thing.... even if it's possible to reduce the rate of injury and fatality to below what a human driver can accomplish, a FSD car and human will kill or injure different people in different circumstances. Human drivers die when they stop paying attention - which computers never do. Computers get confused by unusual circumstances that people can deal with easily.
So, is it ethical or moral to put these cars on the road knowing that, for example, there will be an increase in pedestrian deaths, even if offset by double that in reduction of highway fatalities? The lawyers haven't even begun to sharpen their weapons. It will be a long while before we see cars with no steering wheels.
Daniel in SD
(supervised)
As a frequent pedestrian and cyclist I hope not! Though a large part of the recent increase in pedestrian deaths is due to the proliferation of suvs in cities. I would bet that the Model 3 is one of safest cars to get hit by.
It will take atleast 20 years before we get there. All the existing cars and most of the cars beeing sold the next 5+ years will not have FSD. So people must be able to drive them themselves.
And then you can have manual driving with the FSD running in parallell stopping you from killing yourself or others.
Similar threads
- Article
