3 ms·

But we already let learner drivers try things out on the road, and their supervisors can’t even take control. Human drivers crash and make mistakes too… where i

But we already let learner drivers try things out on the road, and their supervisors can’t even take control. Human drivers crash and make mistakes too… where is the uproar for allowing them to drive? Where as Tesla only allows fully licensed drivers and until recently only pretty good ones at that. I think you’re just regurgitating the standard anti-Tesla talking points… and avoided the questions on the fact Mercedes hasn’t tested this in actual streets and has just rolled it out and while it’s admitting liability isn’t that just as bad as what Tesla is doing? Worse in fact as there no bar for getting this feature, and it doesn’t get shut off for misuse (which FSD beta does). Given all the approaches (with a hope of working) are using AI - how do you get the data to train the AI? This isn’t a Tesla problem, this is an AI training problem. Tesla already makes heavy use of synthetic data, and using real world human drivers to collect training material. How would you do it?

TheAlchemist 3y ago
I don't know the US, but in Europe supervisors definitely can take control - at least for acceleration / breaking. I also think Tesla's approach is the most efficient one, however I do have a problem with their advertising, which is misleading to say the least. I think we kind of all agree it's a beta version, and they 'develop it in the open'. But it's been at least 5 years they say it's pretty much FSD - their first video where they say 'the driver is there only for legal reasons' was in 2016 ! Same goes for accidents data - they don't publish raw data, that would allow some fair comparisons. Instead they publish bullshit charts which are extremely biased.
- kwhitefoot 3y ago
  When I took my sons out to practice driving I most certainly could not take control.
- onethought 3y ago
  But no one publishes detailed data. Every time I hear anti Tesla/Elon chatter, Tesla is always being held to a higher standard than other companies. FSD naming, nvidia call their Chauffeur, which is way more misleading to me. You have Blue Cruise which means nothing. Audi don’t call theirs anything and just talk about the features which I think is your preference. I don’t believe anyone has bought a Tesla not understanding exactly what FSD does and don’t do, are people being misled?
  - TheAlchemist 3y ago
    Uh, you are joking here right ? Tesla uses extremely deceiving marketing (and same goes for investor relations) regarding a lot of their products lately - FSD being one of them. It's called Full Self Driving... Is that misleading ? How do you interpret this name ? Are people being misled ? I don't know. But, last year California did pass a law explicitly aimed at Tesla bullshit. Again, I do think, from society perspective, it's rational thing to do - to 'test in the open' - especially that there are not that much accidents (although the numbers are much worse that Tesla would like people to believe - when you compare apples to apples, there are more accidents per mile with Tesla FSD than with human drivers). What bothers me is their aggressive and misleading marketing, and the fact that they get away with something other companies do not.
justapassenger 3y ago
> But we already let learner drivers try things out on the road, and their supervisors can’t even take control. Human drivers crash and make mistakes too… where is the uproar for allowing them to drive? Drinking water supply also gets contaminated all the time. So I assume you won’t have any issues if I start pouring my experimental medicine into your municipal water supply, right? How else can I get large scale data to test it and develop? Remember, cancer kills millions of people per year! I hope you are not against progress to fight it! > I think you’re just regurgitating the standard anti-Tesla talking points… and avoided the questions on the fact Mercedes hasn’t tested this in actual streets and has just rolled it out and while it’s admitting liability isn’t that just as bad as what Tesla is doing? What makes you think that Mercedes didn’t test on the streets? Just because there were no army of YouTubers, not trained in any way, with no relation to the company, making videos? > Given all the approaches (with a hope of working) are using AI - how do you get the data to train the AI? This isn’t a Tesla problem, this is an AI training problem. Yeah, if you think it’s an AI problem, not how to build and test safety critical systems that’s all I need to know about your understanding of the problem domain.
- onethought 3y ago
  You are clearly trolling rather than trying to understand different points of view. I know it can be fun to be in the pro or anti Tesla camp and run around responding to anyone pro Tesla with your small set of points. But it didn’t progress anyone’s understanding of things. You don’t learn anything and everyone dismisses your views as zealotry. Why not try to engage with the points? Has Mercedes tested this? What evidence do you have for that? If I’m wrong on the AI training problem, how would you collect and validate data without real world feedback? (Ie as you put it “testing in prod). Do other companies do this right?
  - justapassenger 3y ago
    I’m using sarcasm, sure. But do you think it’s unwarranted when I’m against arguments how only person driving the car could be affected by FSD or that self driving is an AI problem? I assume you either don’t argue in good faith, or have understanding of the problem area from following Musk’s and his influencer’s tweets. But I agree with you, it’s not productive. I’ll take my “small set of points” about such an irrelevant things like safety engineering (and no, it’s not a just noun, it’s an engineering discipline) and disengage here. Just to address one thing tho - YouTubers making videos is and Tesla releasing it to more people to recognize deferred revenue is not testing. And you can see from constant regressions that it doesn’t work as testing. And before you throw “well, that’s just how AI works”, maybe you’ll make a connection on how self driving isn’t an AI problem. It’s just one of the many many tools used to solve it.

TheAlchemist 3y ago

I don't know the US, but in Europe supervisors definitely can take control - at least for acceleration / breaking. I also think Tesla's approach is the most efficient one, however I do have a problem with their advertising, which is misleading to say the least. I think we kind of all agree it's a beta version, and they 'develop it in the open'. But it's been at least 5 years they say it's pretty much FSD - their first video where they say 'the driver is there only for legal reasons' was in 2016 ! Same goes for accidents data - they don't publish raw data, that would allow some fair comparisons. Instead they publish bullshit charts which are extremely biased.

kwhitefoot 3y ago
When I took my sons out to practice driving I most certainly could not take control.
onethought 3y ago
But no one publishes detailed data. Every time I hear anti Tesla/Elon chatter, Tesla is always being held to a higher standard than other companies. FSD naming, nvidia call their Chauffeur, which is way more misleading to me. You have Blue Cruise which means nothing. Audi don’t call theirs anything and just talk about the features which I think is your preference. I don’t believe anyone has bought a Tesla not understanding exactly what FSD does and don’t do, are people being misled?
- TheAlchemist 3y ago
  Uh, you are joking here right ? Tesla uses extremely deceiving marketing (and same goes for investor relations) regarding a lot of their products lately - FSD being one of them. It's called Full Self Driving... Is that misleading ? How do you interpret this name ? Are people being misled ? I don't know. But, last year California did pass a law explicitly aimed at Tesla bullshit. Again, I do think, from society perspective, it's rational thing to do - to 'test in the open' - especially that there are not that much accidents (although the numbers are much worse that Tesla would like people to believe - when you compare apples to apples, there are more accidents per mile with Tesla FSD than with human drivers). What bothers me is their aggressive and misleading marketing, and the fact that they get away with something other companies do not.

justapassenger 3y ago

> But we already let learner drivers try things out on the road, and their supervisors can’t even take control. Human drivers crash and make mistakes too… where is the uproar for allowing them to drive? Drinking water supply also gets contaminated all the time. So I assume you won’t have any issues if I start pouring my experimental medicine into your municipal water supply, right? How else can I get large scale data to test it and develop? Remember, cancer kills millions of people per year! I hope you are not against progress to fight it! > I think you’re just regurgitating the standard anti-Tesla talking points… and avoided the questions on the fact Mercedes hasn’t tested this in actual streets and has just rolled it out and while it’s admitting liability isn’t that just as bad as what Tesla is doing? What makes you think that Mercedes didn’t test on the streets? Just because there were no army of YouTubers, not trained in any way, with no relation to the company, making videos? > Given all the approaches (with a hope of working) are using AI - how do you get the data to train the AI? This isn’t a Tesla problem, this is an AI training problem. Yeah, if you think it’s an AI problem, not how to build and test safety critical systems that’s all I need to know about your understanding of the problem domain.

onethought 3y ago
You are clearly trolling rather than trying to understand different points of view. I know it can be fun to be in the pro or anti Tesla camp and run around responding to anyone pro Tesla with your small set of points. But it didn’t progress anyone’s understanding of things. You don’t learn anything and everyone dismisses your views as zealotry. Why not try to engage with the points? Has Mercedes tested this? What evidence do you have for that? If I’m wrong on the AI training problem, how would you collect and validate data without real world feedback? (Ie as you put it “testing in prod). Do other companies do this right?
- justapassenger 3y ago
  I’m using sarcasm, sure. But do you think it’s unwarranted when I’m against arguments how only person driving the car could be affected by FSD or that self driving is an AI problem? I assume you either don’t argue in good faith, or have understanding of the problem area from following Musk’s and his influencer’s tweets. But I agree with you, it’s not productive. I’ll take my “small set of points” about such an irrelevant things like safety engineering (and no, it’s not a just noun, it’s an engineering discipline) and disengage here. Just to address one thing tho - YouTubers making videos is and Tesla releasing it to more people to recognize deferred revenue is not testing. And you can see from constant regressions that it doesn’t work as testing. And before you throw “well, that’s just how AI works”, maybe you’ll make a connection on how self driving isn’t an AI problem. It’s just one of the many many tools used to solve it.