Agree with that except your best guess of 10 years. And not because I think the cars are going to get better fast (it will be asymptotic) .. but because, on average, humans are such crummy drivers and present a very low bar for the car to better. Sure, we can all think of "stump the car" scenarios, but at the end of the day most driver caused accidents are the result of carelessness and/or inattention, something the car doesn't suffer from. So yes, we will get rid of most of the minor accidents, and many of the major ones. But just occasionally the car will do something brain-dead and kill someone. But even seat belts and air bags so that once in a while.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Car Progress
- Thread starter stopcrazypp
- Start date
-
- Tags
- Autonomous Vehicles
Agree with that except your best guess of 10 years. And not because I think the cars are going to get better fast (it will be asymptotic) .. but because, on average, humans are such crummy drivers and present a very low bar for the car to better. Sure, we can all think of "stump the car" scenarios, but at the end of the day most driver caused accidents are the result of carelessness and/or inattention, something the car doesn't suffer from. So yes, we will get rid of most of the minor accidents, and many of the major ones. But just occasionally the car will do something brain-dead and kill someone. But even seat belts and air bags so that once in a while.
One of AI DRIVR's beta10 vid has him fast-forwarding because things are getting boring. I figure we'll know we're there when self-driving becomes like watching paint dry.
boring if you ignore the times the car screwed up and he let it happen since apparently his fun is more important than the comfort of drivers around him who never signed up to be a part of a beta test. One example, Dude let his car go at a stop sign at 13:05 when he didn’t have the right of way and a car was already turning left in front of him. The other car was forced to hit the brakes and let his car go, and the dude didn’t even report it as a problem.
If he wants to experiment with how the car interacts with his wife walking along the side of the road or in the road or across the road with no cars around, have at it! They’re adults and they signed up to be a part of the beta and know what they’re getting into and know what risks are involved. When he lets his car do dumb *sugar* with other drivers on the road that’s just stupid and dangerous and at the very least is an annoyance and a source of discomfort for other drivers on the road who have no idea they’ve been roped into a beta test and just see an apparently incompetent/dangerous Tesla driver, especially when he isn’t even reporting these issues. We know Tesla tracks disengagements, hitting the accelerator, and situations reported by hitting the camera icon so they can review those for improvement. If this dude isn’t doing any of them even when the car does something wrong, how is Tesla supposed to know the car screwed something up and they need to review and fix it?
mikes_fsd
Banned
It seems that not all camera NN's are stitched in yet.
If I was to guess it would be the fender cams and the rear camera, the transitions from cars passing are still jumpy on the videos.
https://twitter.com/elonmusk/status/1353667962213953536
Love that last bit "Critically, however, this does not require a hardware change to cars in field."
If I was to guess it would be the fender cams and the rear camera, the transitions from cars passing are still jumpy on the videos.
https://twitter.com/elonmusk/status/1353667962213953536
Love that last bit "Critically, however, this does not require a hardware change to cars in field."
You're REALLY going out of your way to insult AI DRVR.
I think he showed a few interactions with pedestrians which appear that the car is tracking and anticipating their movements. Nothing seems like it was particularly dangerous.
Tesla has a relatively small number of drivers because they're checking everything. I think the beta testers have reported that their cars are transferring a lot of data with the mothership. Beta drivers probably signed-up to post their videos on YouTube with comments enabled where I'm sure there are lot of vocal people that are determined to see Tesla fail, even more than on TMC.
I think he showed a few interactions with pedestrians which appear that the car is tracking and anticipating their movements. Nothing seems like it was particularly dangerous.
Tesla has a relatively small number of drivers because they're checking everything. I think the beta testers have reported that their cars are transferring a lot of data with the mothership. Beta drivers probably signed-up to post their videos on YouTube with comments enabled where I'm sure there are lot of vocal people that are determined to see Tesla fail, even more than on TMC.
He lets his car impede other drivers on the road basically for fun and YouTube views. That isn’t okay. Confusing and worrying other drivers on the road is a dick move. He’s the one who signed up for a beta, none of the other drivers on the road are interested.
Now THAT is an interesting piece of information.It seems that not all camera NN's are stitched in yet.
If I was to guess it would be the fender cams and the rear camera, the transitions from cars passing are still jumpy on the videos.
https://twitter.com/elonmusk/status/1353667962213953536
View attachment 630520
Love that last bit "Critically, however, this does not require a hardware change to cars in field."
diplomat33
Average guy who loves autonomous vehicles
Oh I see. You have telepathy and can read his mind through your computer. That explains your posts...
Also very convenient that you change topics, don't address any of your prior criticisms, and immediately throw up more straw-men.
Also very convenient that you change topics, don't address any of your prior criticisms, and immediately throw up more straw-men.
Oh I see. You have telepathy and can read his mind through your computer. That explains your posts...
Also very convenient that you change topics, don't address any of your prior criticisms, and immediately throw up more straw-men.
I can literally SEE HIM letting him car impede other drivers. The fact he’s doing it for a Youtube video and mentions his patreons at the start of every video makes it easy to conclude why he’s doing it. No telepathy needed.
ladysbff
Member
Lmao you did the exact same thing to me Mark my old friend. The projection.. it burns..
diplomat33
Average guy who loves autonomous vehicles
The article is a bit long but interesting IMO. He argues that there is no "race to win FSD". There won't be a single "winner". Rather, FSD will develop and get better over time and eventually self-driving cars will be ubiquitous, like elevators, with several companies sharing the market.
"Who will win the self-driving car race? People ask me this all the time, and the question doesn’t make much sense. It assumes there actually is a race, with a clear finish line and one winner. Does anyone remember who “won” trains or ships or aviation? The Wright Brothers didn’t “win” the aviation business when their biplane flew 852 feet in 1903."
Who Will Win The Self-Driving Car Race? The Clues Lie In Elevator History
"Who will win the self-driving car race? People ask me this all the time, and the question doesn’t make much sense. It assumes there actually is a race, with a clear finish line and one winner. Does anyone remember who “won” trains or ships or aviation? The Wright Brothers didn’t “win” the aviation business when their biplane flew 852 feet in 1903."
Who Will Win The Self-Driving Car Race? The Clues Lie In Elevator History
powertoold
Active Member
I haven't read the article, but the FSD race is akin to, "which company will be the first to provide an Uber-esque robotaxi service to the general public?"
But yes, everyone has different opinions about what the race is about.
But yes, everyone has different opinions about what the race is about.
I'd like to see the race to which manufacturer will take liability for accidents first with FSD capability
The first company to offer a sleep-in-the-back self-driving car for sale to the general public in numbers large enough that anybody who wants one can get it without an excessive waiting period will have a tremendous market advantage. Tesla was the first company to sell mass-manufactured electric cars that are not golf carts and is far and away the industry leader. The first SItB FSD will be like that.
We're going to see gradual improvements in driver-assist features that will blend into FSD. But the first company to flip the switch and say "You can sleep in the back" will be hard to beat, unless they screw it up.
The elevator analogy doesn't really fit because safety testing is much more sophisticated and I'm not talking about the process of invention, but about a critical moment in a mature multi-billion-dollar industry. Robotaxis will spread slowly, but there will be one moment and one company that puts actual, real, genuine self-driving cars into private consumers' hands.
You may be right. On the other hand, I don't mind sitting at the steering wheel, ready to take over, particularly not if the car drives nearly perfectly and warns me in time before it gets into trouble. If I can read a newspaper or chat with passengers in a relaxed situation while keeping myself ready to grab the steering wheel, I'll be happy enough.
You may still be right. But perhaps the real breakthrough will not happen when I can sleep in the back, but when I can order my car to come to me without a driver or to collect food from my favorite restaurant.
I would like to be able to "summon" the car from a distance, or dispatch my car to a location. As a pilot, it would be nice to dispatch your car to an airport and have it waiting there for your arrival, and then send it back home when you are done! This is autonomous!!
diplomat33
Average guy who loves autonomous vehicles
There is no doubt that the first SitB L5 consumer car will be a huge game changer. And it will happen some day. But it is unlikely that one company alone will do it en masse before everybody else. It is more likely that we will see the spread of L4 robotaxis. We will see the spread of advanced driver assist on consumer cars. We will see L3 on consumer cars. And several AV companies will probably achieve SitB L5 consumer cars at roughly the same time. So consumers will have a choice. That is why the article suggests that there won't be just 1 FSD winner, there will be multiple FSD winners that share the pot.
+1 .. this is going to be one of those mfg X gets A right, but B wrong, while mfg Y get B right and A wrong. Eventually, everyone gradually converges on the feature set we all want, and then there is a rush to the bottom as self-driving ceases to be a product differentiator. But that's 20 years down the road probably (just a swag).
Similar threads
- Article
