7 ms·

I met someone recently who told me they SLEPT on their way to work in their Tesla. On purpose. They bought the car specifically because they had a very long 2.5

I met someone recently who told me they SLEPT on their way to work in their Tesla. On purpose. They bought the car specifically because they had a very long 2.5 hour commute. The marketing for this stuff is absolutely liable and as much as I like what Tesla has done for electric cars in general we are one big, massive accident away from this all absolutely blowing up.

twblalock 4y ago
> I met someone recently who told me they SLEPT on their way to work in their Tesla. On purpose. They bought the car specifically because they had a very long 2.5 hour commute. Think about that. If someone can literally sleep while the car is driving itself, and it doesn't crash, then the self-driving feature actually works really well. I mean, try doing that in a Chevy.
- Veserv 4y ago
  No. Humans average something like 250,000 miles per insurance-worthy event. Even if we assumed a average speed of 80 mph, which is absurdly high, that is ~3,000 hours between accidents. That system would need to make the interaction-less roundtrip commute 600 times, 2.5 work years, without even a single minor accident for the unassisted system to even be comparable to the average driver.
  - twblalock 4y ago
    Why are you comparing it to a human driver? Compare it to other self-driving cars and you’ll see that it is actually one of the better systems on the market.
    - wewtyflakes 4y ago
      Because people buy this feature because they don't want to drive (as a human); what does it matter how it compares to other self driving systems?
      - twblalock 4y ago
        Because it undermines the argument made in the "criminal probe". Tesla is doing self-driving as well as it can possibly be done with today's technology. If it could be done better, it would be, by one of their competitors, but they can't beat it. It is in fact so good that people can sleep in their cars and not crash -- and I wouldn't bet that the guy who sleeps in the car only did it once! If you are going to compare it to human drivers, it also matters which humans you compare it to. Every time I drive I see people doing things far stupider than what a Tesla would do. Driving skill, and judgment, is so unevenly distributed in the human population that it would not be difficult to find a large number of humans who underperform the Tesla system. It would be weird to punish Tesla for not being perfect. It has delivered self driving features that are already better than a lot of the humans on the roads today, and better than all of its competitors. Let's go a bit deeper into your argument: > Because people buy this feature because they don't want to drive (as a human); what does it matter how it compares to other self driving systems? Ok, pretend you are someone who is looking to buy a self-driving car because you don't want to do the driving. Wouldn't you want to buy one that works well? It's a matter of life and death, after all. So you would do some research to find out who had the best systems. If you did that you would find out that Tesla's system, despite all the criticism it gets, is one of the few on the market that is worth buying.
        cactusplant7374 4y ago
        > It is in fact so good that people can sleep in their cars and not crash -- and I wouldn't bet that the guy who sleeps in the car only did it once! This thinking is absolutely going to get someone killed. It’s also why Tesla faces so much criticism. Any improvement under 99.999% is unremarkable because the end product is unusable. No one is going to buy a car that crashes once a year. Disengagements should be unheard of if Elon wants to continue selling FSD.
        chologrande 4y ago
        Any improvement _over_ 99.999% is impossible. Saying something like this is completely detached from reality. People _do_ buy cars that crash once a year. All the time. And often the defects in the system aren't even related to self driving functionality. Further, most drivers who get in a car have records far worse than that.
    - Veserv 4y ago
      Are we really going there? About 40,000 people die per year in the US from traffic fatalities. There are about 200,000,000 people licensed to drive in the US. So, about 1 in 5,000 drivers die per year. Tesla has released FSD Beta to ~160,000 drivers. So, to first order, we can approximate that as ~32 Tesla drivers will die a year due to a traffic fatality if they constitute a representative sample (they do not, but we are doing some quick estimates here). So, a system that is used as a full-time, unassisted replacement for their human drivers will result in, on average, one extra person dying per year per 3% worse it is than a human driver when applied to a mere 0.1% of the US driving population. If applied over the entire US driving population a system that is 3% worse applied to all driving interactions would result in ~1,200 extra people dying per year. At the scales being discussed, extremely small differences applied over the entire domain result in very serious risks to human life. It is not okay to knowingly sacrifice people at the altar of moving fast and breaking things so that Elon Musk can make another couple billion dollars. Also, it is not even close to being one of the better systems from a safety perspective. It is something like 400-800x worse than Waymo and 2,000-4,000x worse than Cruise [1]. It is so far behind so many other companies it is fair to say that they are not even in the race let alone a frontrunner. [1] https://news.ycombinator.com/item?id=33227450 https://news.ycombinator.com/item?id=33227450
      - FeepingCreature 4y ago
        Though to be fair, the faster the system crosses the 3% mark the faster that 40,000 number can come down.
        Veserv 4y ago
        They are not within 3% or even close. They are 2,000x worse than Cruise which is still at least 8x worse than human drivers. Being generous they are 16,000x (1,600,000%) worse at unassisted, intervention-less driving than humans. To put that into perspective, if every car in the US had Tesla FSD Beta and they were all using it as a fully autonomous system without babysitting, it would average 1,750,000 deaths per day and everybody in the US would be dead in 6 months. Within a week it would kill more people than have ever died from traffic accidents in the US. We are literally talking 1 deca-Hiroshima per day of badness. The only saving grace of this whole thing is that hardly anybody is insane enough to use it unattended more than a few times so we do not see the sheer catastrophe that would occur if it was actually used as advertised.
    - thwayunion 4y ago
      > Compare it to other self-driving cars No one else sells self-driving cars to consumers because they know they cannot. "Being unable to do something and not saying you can do it" "Being unable to do something and saying you can do it" Spot the difference?
    - TheCoelacanth 4y ago
      As someone who could potentially be killed by one of Elon's homicidemobiles, I would very much prefer if vehicles that are more deadly than those driven by humans would not drive in places where they could potentially kill me.
      - twblalock 4y ago
        You’re a lot more likely to be killed by a young idiot in a Dodge Charger.
- dangrossman 4y ago
  > I mean, try doing that in a Chevy. I was under the impression Chevy SuperCruise outperformed Tesla FSD in most ways, other than being bound to mapped roads.
  - modeless 4y ago
    SuperCruise and FSD currently don't even work on any of the same roads. FSD is for non-freeway roads, and SuperCruise only works on (some) freeways. Autopilot is Tesla's system that works on freeways. (Tesla does plan to merge Autopilot and FSD eventually, and SuperCruise will expand to some non-freeway roads, but neither has happened yet). SuperCruise is only rated higher than Autopilot when people use rating systems that dock Autopilot for having fewer driver monitoring features, i.e. nanny features that scold you for not paying enough attention. I haven't seen any evidence that the actual driving part of SuperCruise or any other competitor is better than Autopilot.
    - dham 4y ago
      OpenPilot is probably the best or second best on the market from my usage of other systems.
    - asdff 4y ago
      What's going on with this then? FSD on freeways 2 years ago: https://www.youtube.com/watch?v=b8aw_nCGIJw https://www.youtube.com/watch?v=b8aw_nCGIJw
      - modeless 4y ago
        That's Autopilot on freeways, not FSD
        asdff 4y ago
        Its merging lanes in the video though, isn't that FSD? I thought autopilot held your lane.
JohnJamesRambo 4y ago
I’ll never understand anyone that has this much faith in technology. Too many blue screens in my formative years. Not do I understand anyone trusting their Tesla this much. Even the cruise control freaks out sometimes and slams on the brakes from phantom apparitions. I don’t even trust the base autosteer feature going down the highway.
- steve_adams_86 4y ago
  > Too many blue screens in my formative years. Those blue screens occurred in dramatically less complex and better understood software, too. I'm not saying we don't understand how FSD works, or the underlying technology. It's not a complete black box. I believe there was even a proof recently showing that these things are essentially massive decision trees and they can be deterministically unravelled. But this is otherwise totally negligent from where I'm standing.
  - pnt12 4y ago
    A neural network doesn't panic, so there is no blue screen, only bad decisions. I think that's even worse, as that would be a strong signal to take over driving. This might be mitigated if the result includes a confidence percentage, ie panic if confidence below threshold, but even then I'm not sure how reliable this metric is.
    - steve_adams_86 4y ago
      Absolutely. I worry that the decision tree is so complex that when we do traverse it to understand errors, we could easily draw the wrong conclusions anyway. When we need to audit something that is more complex than we can really comprehend, can we come to reliably accurate conclusions? If we need tools to do this for us, can we be certain they’re accurate as well? All of it feels very Icarus-like. I’d be content if we tested this stuff out in isolation for another 20 years before throwing it on public streets.
mlindner 4y ago
> I met someone recently who told me they SLEPT on their way to work in their Tesla. He'd have to work around and bypass several safety features that prevent you from doing this. You cannot do this out of the box. You have to work hard to get it to do it. I take any claims of people doing this with a large grain of salt.
- willio58 4y ago
  This is common in the community though, you literally just have to use a certain weight on the wheel
  - mlindner 4y ago
    They also have awareness detection, if you're on the beta, as well. A simple weight is not enough.
    - dawnerd 4y ago
      Ways around this too sadly.
gnicholas 4y ago
Yeah, I was not thrilled when a friend gave me a ride in a Tesla and was nodding off as we sped down 101. I kept trying to ask him questions to keep him awake, with limited success.
- Balgair 4y ago
  Hey, I don't know if you need to hear this, but: I think your friend isn't very good at being a friend back to you. You may need to re-evaluate.
  - germinalphrase 4y ago
    Second.
  - gnicholas 4y ago
    Thanks for your concern; I was using the term loosely. This is a friend-of-a-friend, who was giving me a ride. Next time I'll drive myself!
    - Balgair 4y ago
      Glad to hear all is okay!
willio58 4y ago
So unsafe. I work with someone who’s very pro-Tesla and she let AP beta drive us around the Seattle area a few times when I visited a while ago. We got into 2 near-accidents. Firstly, the car ran over a cone on the freeway going 65mph+. Second, it almost turned into oncoming traffic out of nowhere driving in a residential/commercial area.
bergenty 4y ago
It works really well in any place that isn’t a dense city. Out here in the suburbs it’s pretty fantastic.
- zhoujianfu 4y ago
  Out here in west LA it’s a joke! But I have thought to myself, “this thing would work pretty well in rural/suburban settings.”
sngz 4y ago
I drove next to a guy on the highway that was sleeping in his tesla. his head was cocked backwards with his mouth wide open probably snoring. we were probably going 70mph.
LeanderK 4y ago
> very long 2.5 hour commute imagine doing nothing else besides working and commenting. I can't. It sounds like a recipe for a lonely life.
smaryjerry 4y ago
I don’t think it has to be a bad thing. It be an interesting statistical problem. What is worse, a person who drives sleepy for 5 hours each day or a person who gives up control to a computer for 5 hours each day. Point being we are attempting to compare FSD to complete perfect driving, when the comparison should be with current drivers. If it is an improvement or reduction in crashes why not use it? Maybe it’s because we aren’t sure who to blame in case something goes wrong? I know back seat drivers that get uncomfortable as a passenger a lot even with a driver with zero accidents and would rather drive themselves, yet have crashed several times themselves at their fault. They misjudge their own abilities and the situation, a need for certain aggressive driving. I think many are misjudging the average drivers ability without self driving and we could be causing a lot of loss of life by not taking a statistically better option out of human hubris.
- ycombobreaker 4y ago
  This is not a valid comparison, though. Rather, consider the extrapolation: "drive for 20 hours or let a computer drive for 20 hours?". (edit: the extrapolation is intended to highlight the absurdity of the premise.) The correct answer is to mitigate the risk with breaks or safer alternative transport such as a hired driver. It's not a closed system with only two possible solutions. If the solution is not tenable, the plan should be aborted. Not rammed through via abuse of technology.
  - smaryjerry 4y ago
    I didn’t say it was a closed system. I’m just saying one might be better than the other, I don’t know how you do that other than compare the possibilities. If everyone had a private driver then yea that options is also better but we live in a human world not a perfect world.
    - ycombobreaker 4y ago
      You implied that the only solution to an unsafe decision was a similarly unsafe decision. My argument is that the real solution would involve some other compromise. If this specific scenario was a work commute, for instance: - move closer to work - secure lodging to amortize the commute over more days - arrange a carpool - work remotely to reduce the cost of any of the above options - find a different job. (edit for list format)

twblalock 4y ago

> I met someone recently who told me they SLEPT on their way to work in their Tesla. On purpose. They bought the car specifically because they had a very long 2.5 hour commute. Think about that. If someone can literally sleep while the car is driving itself, and it doesn't crash, then the self-driving feature actually works really well. I mean, try doing that in a Chevy.

Veserv 4y ago
No. Humans average something like 250,000 miles per insurance-worthy event. Even if we assumed a average speed of 80 mph, which is absurdly high, that is ~3,000 hours between accidents. That system would need to make the interaction-less roundtrip commute 600 times, 2.5 work years, without even a single minor accident for the unassisted system to even be comparable to the average driver.
- twblalock 4y ago
  Why are you comparing it to a human driver? Compare it to other self-driving cars and you’ll see that it is actually one of the better systems on the market.
  - wewtyflakes 4y ago
    Because people buy this feature because they don't want to drive (as a human); what does it matter how it compares to other self driving systems?
    - twblalock 4y ago
      Because it undermines the argument made in the "criminal probe". Tesla is doing self-driving as well as it can possibly be done with today's technology. If it could be done better, it would be, by one of their competitors, but they can't beat it. It is in fact so good that people can sleep in their cars and not crash -- and I wouldn't bet that the guy who sleeps in the car only did it once! If you are going to compare it to human drivers, it also matters which humans you compare it to. Every time I drive I see people doing things far stupider than what a Tesla would do. Driving skill, and judgment, is so unevenly distributed in the human population that it would not be difficult to find a large number of humans who underperform the Tesla system. It would be weird to punish Tesla for not being perfect. It has delivered self driving features that are already better than a lot of the humans on the roads today, and better than all of its competitors. Let's go a bit deeper into your argument: > Because people buy this feature because they don't want to drive (as a human); what does it matter how it compares to other self driving systems? Ok, pretend you are someone who is looking to buy a self-driving car because you don't want to do the driving. Wouldn't you want to buy one that works well? It's a matter of life and death, after all. So you would do some research to find out who had the best systems. If you did that you would find out that Tesla's system, despite all the criticism it gets, is one of the few on the market that is worth buying.
      - cactusplant7374 4y ago
        > It is in fact so good that people can sleep in their cars and not crash -- and I wouldn't bet that the guy who sleeps in the car only did it once! This thinking is absolutely going to get someone killed. It’s also why Tesla faces so much criticism. Any improvement under 99.999% is unremarkable because the end product is unusable. No one is going to buy a car that crashes once a year. Disengagements should be unheard of if Elon wants to continue selling FSD.
        chologrande 4y ago
        Any improvement _over_ 99.999% is impossible. Saying something like this is completely detached from reality. People _do_ buy cars that crash once a year. All the time. And often the defects in the system aren't even related to self driving functionality. Further, most drivers who get in a car have records far worse than that.
  - Veserv 4y ago
    Are we really going there? About 40,000 people die per year in the US from traffic fatalities. There are about 200,000,000 people licensed to drive in the US. So, about 1 in 5,000 drivers die per year. Tesla has released FSD Beta to ~160,000 drivers. So, to first order, we can approximate that as ~32 Tesla drivers will die a year due to a traffic fatality if they constitute a representative sample (they do not, but we are doing some quick estimates here). So, a system that is used as a full-time, unassisted replacement for their human drivers will result in, on average, one extra person dying per year per 3% worse it is than a human driver when applied to a mere 0.1% of the US driving population. If applied over the entire US driving population a system that is 3% worse applied to all driving interactions would result in ~1,200 extra people dying per year. At the scales being discussed, extremely small differences applied over the entire domain result in very serious risks to human life. It is not okay to knowingly sacrifice people at the altar of moving fast and breaking things so that Elon Musk can make another couple billion dollars. Also, it is not even close to being one of the better systems from a safety perspective. It is something like 400-800x worse than Waymo and 2,000-4,000x worse than Cruise [1]. It is so far behind so many other companies it is fair to say that they are not even in the race let alone a frontrunner. [1] https://news.ycombinator.com/item?id=33227450 https://news.ycombinator.com/item?id=33227450
    - FeepingCreature 4y ago
      Though to be fair, the faster the system crosses the 3% mark the faster that 40,000 number can come down.
      - Veserv 4y ago
        They are not within 3% or even close. They are 2,000x worse than Cruise which is still at least 8x worse than human drivers. Being generous they are 16,000x (1,600,000%) worse at unassisted, intervention-less driving than humans. To put that into perspective, if every car in the US had Tesla FSD Beta and they were all using it as a fully autonomous system without babysitting, it would average 1,750,000 deaths per day and everybody in the US would be dead in 6 months. Within a week it would kill more people than have ever died from traffic accidents in the US. We are literally talking 1 deca-Hiroshima per day of badness. The only saving grace of this whole thing is that hardly anybody is insane enough to use it unattended more than a few times so we do not see the sheer catastrophe that would occur if it was actually used as advertised.
  - thwayunion 4y ago
    > Compare it to other self-driving cars No one else sells self-driving cars to consumers because they know they cannot. "Being unable to do something and not saying you can do it" "Being unable to do something and saying you can do it" Spot the difference?
  - TheCoelacanth 4y ago
    As someone who could potentially be killed by one of Elon's homicidemobiles, I would very much prefer if vehicles that are more deadly than those driven by humans would not drive in places where they could potentially kill me.
    - twblalock 4y ago
      You’re a lot more likely to be killed by a young idiot in a Dodge Charger.
dangrossman 4y ago
> I mean, try doing that in a Chevy. I was under the impression Chevy SuperCruise outperformed Tesla FSD in most ways, other than being bound to mapped roads.
- modeless 4y ago
  SuperCruise and FSD currently don't even work on any of the same roads. FSD is for non-freeway roads, and SuperCruise only works on (some) freeways. Autopilot is Tesla's system that works on freeways. (Tesla does plan to merge Autopilot and FSD eventually, and SuperCruise will expand to some non-freeway roads, but neither has happened yet). SuperCruise is only rated higher than Autopilot when people use rating systems that dock Autopilot for having fewer driver monitoring features, i.e. nanny features that scold you for not paying enough attention. I haven't seen any evidence that the actual driving part of SuperCruise or any other competitor is better than Autopilot.
  - dham 4y ago
    OpenPilot is probably the best or second best on the market from my usage of other systems.
  - asdff 4y ago
    What's going on with this then? FSD on freeways 2 years ago: https://www.youtube.com/watch?v=b8aw_nCGIJw https://www.youtube.com/watch?v=b8aw_nCGIJw
    - modeless 4y ago
      That's Autopilot on freeways, not FSD
      - asdff 4y ago
        Its merging lanes in the video though, isn't that FSD? I thought autopilot held your lane.

JohnJamesRambo 4y ago

I’ll never understand anyone that has this much faith in technology. Too many blue screens in my formative years. Not do I understand anyone trusting their Tesla this much. Even the cruise control freaks out sometimes and slams on the brakes from phantom apparitions. I don’t even trust the base autosteer feature going down the highway.

steve_adams_86 4y ago
> Too many blue screens in my formative years. Those blue screens occurred in dramatically less complex and better understood software, too. I'm not saying we don't understand how FSD works, or the underlying technology. It's not a complete black box. I believe there was even a proof recently showing that these things are essentially massive decision trees and they can be deterministically unravelled. But this is otherwise totally negligent from where I'm standing.
- pnt12 4y ago
  A neural network doesn't panic, so there is no blue screen, only bad decisions. I think that's even worse, as that would be a strong signal to take over driving. This might be mitigated if the result includes a confidence percentage, ie panic if confidence below threshold, but even then I'm not sure how reliable this metric is.
  - steve_adams_86 4y ago
    Absolutely. I worry that the decision tree is so complex that when we do traverse it to understand errors, we could easily draw the wrong conclusions anyway. When we need to audit something that is more complex than we can really comprehend, can we come to reliably accurate conclusions? If we need tools to do this for us, can we be certain they’re accurate as well? All of it feels very Icarus-like. I’d be content if we tested this stuff out in isolation for another 20 years before throwing it on public streets.

mlindner 4y ago

> I met someone recently who told me they SLEPT on their way to work in their Tesla. He'd have to work around and bypass several safety features that prevent you from doing this. You cannot do this out of the box. You have to work hard to get it to do it. I take any claims of people doing this with a large grain of salt.

willio58 4y ago
This is common in the community though, you literally just have to use a certain weight on the wheel
- mlindner 4y ago
  They also have awareness detection, if you're on the beta, as well. A simple weight is not enough.
  - dawnerd 4y ago
    Ways around this too sadly.

gnicholas 4y ago

Yeah, I was not thrilled when a friend gave me a ride in a Tesla and was nodding off as we sped down 101. I kept trying to ask him questions to keep him awake, with limited success.

Balgair 4y ago
Hey, I don't know if you need to hear this, but: I think your friend isn't very good at being a friend back to you. You may need to re-evaluate.
- germinalphrase 4y ago
  Second.
- gnicholas 4y ago
  Thanks for your concern; I was using the term loosely. This is a friend-of-a-friend, who was giving me a ride. Next time I'll drive myself!
  - Balgair 4y ago
    Glad to hear all is okay!

willio58 4y ago

So unsafe. I work with someone who’s very pro-Tesla and she let AP beta drive us around the Seattle area a few times when I visited a while ago. We got into 2 near-accidents. Firstly, the car ran over a cone on the freeway going 65mph+. Second, it almost turned into oncoming traffic out of nowhere driving in a residential/commercial area.

bergenty 4y ago

It works really well in any place that isn’t a dense city. Out here in the suburbs it’s pretty fantastic.

zhoujianfu 4y ago
Out here in west LA it’s a joke! But I have thought to myself, “this thing would work pretty well in rural/suburban settings.”

sngz 4y ago

I drove next to a guy on the highway that was sleeping in his tesla. his head was cocked backwards with his mouth wide open probably snoring. we were probably going 70mph.

LeanderK 4y ago

> very long 2.5 hour commute imagine doing nothing else besides working and commenting. I can't. It sounds like a recipe for a lonely life.

smaryjerry 4y ago

I don’t think it has to be a bad thing. It be an interesting statistical problem. What is worse, a person who drives sleepy for 5 hours each day or a person who gives up control to a computer for 5 hours each day. Point being we are attempting to compare FSD to complete perfect driving, when the comparison should be with current drivers. If it is an improvement or reduction in crashes why not use it? Maybe it’s because we aren’t sure who to blame in case something goes wrong? I know back seat drivers that get uncomfortable as a passenger a lot even with a driver with zero accidents and would rather drive themselves, yet have crashed several times themselves at their fault. They misjudge their own abilities and the situation, a need for certain aggressive driving. I think many are misjudging the average drivers ability without self driving and we could be causing a lot of loss of life by not taking a statistically better option out of human hubris.

ycombobreaker 4y ago
This is not a valid comparison, though. Rather, consider the extrapolation: "drive for 20 hours or let a computer drive for 20 hours?". (edit: the extrapolation is intended to highlight the absurdity of the premise.) The correct answer is to mitigate the risk with breaks or safer alternative transport such as a hired driver. It's not a closed system with only two possible solutions. If the solution is not tenable, the plan should be aborted. Not rammed through via abuse of technology.
- smaryjerry 4y ago
  I didn’t say it was a closed system. I’m just saying one might be better than the other, I don’t know how you do that other than compare the possibilities. If everyone had a private driver then yea that options is also better but we live in a human world not a perfect world.
  - ycombobreaker 4y ago
    You implied that the only solution to an unsafe decision was a similarly unsafe decision. My argument is that the real solution would involve some other compromise. If this specific scenario was a work commute, for instance: - move closer to work - secure lodging to amortize the commute over more days - arrange a carpool - work remotely to reduce the cost of any of the above options - find a different job. (edit for list format)