11 ms·

Having finally had a chance to test-"drive" an advanced autopilot Tesla last weekend, I can confidently say at the current level of that tech I would pay less t

Having finally had a chance to test-"drive" an advanced autopilot Tesla last weekend, I can confidently say at the current level of that tech I would pay less than nothing. At best (cruising down a well-marked highway) it was only slightly worse than useless, taking a different line in the lane than I considered optimal, leaving me feeling a background level of apprehension. From there it ranged through frustrating, as it took forever to do anything, to jarring, as it would for example suddenly slam on the brakes because it just noticed it was passing a parked car. If I had a regular commute and self-driving was considerably beyond human competence I'd love to have it, but I don't see either one happening any time soon!

mlyle 5y ago
Tesla's status on this is really controversial. Some people think that Tesla's sheer data advantage will put them way ahead soon and that it's imminent. On the other hand, Waymo is doing pretty well on most trips (admittedly within a small geofenced zone that has friendly weather -- you can watch oodles of video from jjricks on real trips here: https://www.youtube.com/watch?v=61AfxOMu3wY https://www.youtube.com/watch?v=61AfxOMu3wY )
- tempestn 5y ago
  I actually have watched a number of his videos, and while it is considerably better than the Tesla self driving, it's not something I'd currently pay to experience vs driving myself (or vs an Uber at a similar price point). The driving style is still too awkward and, to me, discomfiting.
  - mlyle 5y ago
    Ah, to me it looks pretty good. And there's reason to believe it'd be better on freeways. I'd definitely pay a few bucks per hour for my car to drive me to Southern California while I read a book.
    - gitanovic 5y ago
      What I don't understand is: why don't you take a train instead? I have never been to US, but whenever I hear about self driving cars I think: why don't people just use public transport if they don't want to drive?
      - gnicholas 5y ago
        Strangely, there are no direct trains from northern CA to southern CA. There is a line from Sacramento down into Santa Clara, and a line from San Luis Obispo down to San Diego. But there's no direct way to get from SF/SJ to LA/OC via train. There are buses, but they are considerably less enjoyable than the Pacific Surfliner line!
      - blhack 5y ago
        It’s too much of a hassle to use public transportation. With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc. Like for instance with my own car instead of a shared train, I can bring my golf clubs, my dog, and my drone with me, and call my friend on the way out of town to see if he’s up for a spontaneous trip to San Diego to see what fun we could get up to there for the weekend. That’s not really possible on a train/airplane. It’s just a totally different experience than arriving to a train station with some limited set of your things, possibly being late, then arriving somewhere else again without a car, and having to return again by doing all of this in reverse. It also means spending time forced to socialize with people you don’t know, and almost definitely in a seat that is less comfortable than the one in your $100k luxury car.
        darkwater 5y ago
        > With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc. Also, you can get blocked in a traffic jam when you don't want.
- tails4e 5y ago
  I did think it was imminent at one point, but the fact the delivery date keeps pushing out shows its far from complete, and is only usable in controlled circumstances. I use it in traffic jams, or motorway driving, and that's it, anywhere else and it's just asking for trouble. It's useful in those limited cases, but it's a shame how many false promises have been issued, I've a lot of respect for what Elon has done, but he's really oversold this
- malozite 5y ago
  I think the data advantage is limited. I used to work in high-frequency equity trading. We would run simulations against tick-by-tick data from real markets. The simulations were of very questionable value. As soon as you trade once in your simulation, from that point onwards the whole run is contaminated. Your trade, if it had actually happened, would have been seen by other participants, who would have changed their strategy, often in complex and unpredictable ways, and the timelines would have diverged rapidly and irreversibly from that point. Data recorded from driving situations is useful for one thing - testing out split-second decision making - stop or accelerate? Steer left or right? This is a limited use case. Most situations don't have critical decisions like this. The ones that do, it's often hard to judge what would have been the right answer. Even accidents - where the post facto desired outcome is "don't have an accident", unfold over several seconds, involving several stages of actors responding to each others' actions. No one can be sure how the driver should have approached the very first interaction - the only one for which the data is meaningful. Actual road testing, and meaningful simulation which isn't just replaying what actually happened is much more useful.
- lm28469 5y ago
  > Waymo Even they admitted fully autonomous vehicles most likely won't be a thing any time soon (maybe ever if we keep our current infrastructure) https://www.cnet.com/news/alphabet-google-waymo-ceo-john-krafcik-autonomous-cars-wont-ever-be-able-to-drive-in-all-conditions/ https://www.cnet.com/news/alphabet-google-waymo-ceo-john-kra...
  - mlyle 5y ago
    I feel like this was misread. The statement appears to say that he doesn't think they will completely displace manual driving of cars for considerable time. And that SAE level 5 arguably expects more than we expect from human drivers (many of us would not be safe driving in whiteout conditions on a windy mountain road).
hughrr 5y ago
So much this. I hope I don't stand on too many toes here. Actually I don't care. My father has a self driving enabled Tesla Model S which he paid a hell of a lot of money for that he doesn't have as he's in his 70s, living entirely off the remains of the house he just sold. He's convinced it'll be any day now that he gets full self driving and that it'll change the world. He tells me he loves the car and phones me daily to tell me about it. But at the end of the day there are a bunch of complaints hiding in his words. This covers everything from poor fitting, paint issues, weird bugs, massive safety and computer issues, lots of hidden costs and general battery life issues. He's hiding the true picture which is not the real experience. Between Tesla marketing and the echo chamber Tesla forums, he's traded his entire cash pile and attention for what I can only describe as a prototype that is a ball and chain which limps between bouts of charging CPR and he can't tell anyone because it might damage his conspicuous virtuous electric car identity. But Tesla will be introducing self driving any day now he tells me. All the time. For the last 3 years... I'm not really a proponent or opponent of electric cars. I don't really care about the things as they don't factor into my life past the utility of a screwdriver. I myself drive a third-hand Citroen with a tuk tuk engine in it which was cheap to buy and run and takes me to the shops and for the odd weekend out. But from my vantage point, I can't see the return of investment that such complex technology can add to a simple problem. All it does is make it more complicated, less safe and less reliable while milking people for cash for promises which may never materialise. If there's anything I've worked out over the last couple of decades in the software industry is that this model really doesn't work - it's the marketing and development model that Microsoft uses. Edit: he literally just phoned me to tell me how wonderful it is again. But it cost him 5% of the battery capacity to cool it down enough that it’s not going to burn his ass off (has sunroof) which means his next journey is going to have to end with a 10 mile detour to sit in a supercharger bay (which where he lives will be busy and he’ll have to wait) and thus probably cost a coffee at the same time.
- ec109685 5y ago
  Obviously you know your father best, but it at least seems from your words it makes him happy. As someone with similar car, you get to know it’s quirks and use it when it adds value.
Rygian 5y ago
What did you test-drive? Regular Tesla auto-pilot, or the limited-access FSD Beta? They are radically different experiences.
- mft_ 5y ago
  This is important. From the original description, my suspicion is it’s the regular (i.e. not advanced beta) “full self driving“ package, which I have on my model 3. Truth is, it’s pretty limited, and you’ve got to learn the types of situations where it will and won’t cope. Through town, situations with parked cars, or weird lane markings, it will struggle. However, the flipside is that on longer journeys – with mostly open roads, it does very well indeed. I recently did a three+ hour drive and the car drove for probably 80% of that time, and without any issues at all. Once you get used to its foibles and therefore understand when you can trust it, it really is a relaxing experience.
  - Dah00n 5y ago
    The FSD beta is way worse. You have to be much more conscious about everything that happens around you than with the "old package". Then you might as well drive yourself. In my experience it's like how it feels being in a car driven by an old relative that clearly shouldn't be driving a car anymore but you're too polite to tell him.
    - ketamine__ 5y ago
      The accidents are going to increase when FSD is released to the public with subscriptions.
- tempestn 5y ago
  I'm not certain since it was just a test drive, and there was no Tesla rep with me - it did say "Beta" on the autopilot screen, and seemed happy to engage on any kind of street, but I don't know if that means it was the full FSD Beta. (I'd be kind of surprised if they just handed people the (virtual) keys to that with no warnings, but Tesla does tend to do surprising things.)
berkes 5y ago
I'm no expert on any of the selfdriving features (I still self-drive my 1984 Volkswagen with tlc) but from what I hear and see, Tesla is amongst the worst when it comes to selfdriving and e.g. Volvo or Volkwagen is way ahead. Their marketing, however, is top notch, so whenever Tesla fails to deliver, people conclude that 'all selfdriving must be crap'. Is there something called 'Law of inhibiting leadership'? In which an underperforming leader of a segment holds back the entire segment?
- terryf 5y ago
  As an owner of a Volvo V90 that has pilot assist or whatever they call it - it's great for highway driving! It has it's quirks but they are easy to get used to and after a few years, I know pretty much exactly where it is extremely reliable and where it will get it wrong. The behaviour is extremely boringly reliable - which is exactly what it should be and doing a few hundrer km on a good highway is now much much more relaxing than before.
lm28469 5y ago
Don't worry Musk said it'll be delivered in two years tops, in 2012 or something. Anyone still believing in fully autonomous cars as they are presented by Musk &co today (the real thing, not a cruise control that only works on California's highways on a sunny day) is 101% delusional. We'll talk about it in 15 years as the "previous AI hype cycle".
lvl100 5y ago
Tesla scammed people into believing in FSD. I actually believed it for a second too. Yet here we are with Elon Musk as one of the richest con of all time.

mlyle 5y ago

Tesla's status on this is really controversial. Some people think that Tesla's sheer data advantage will put them way ahead soon and that it's imminent. On the other hand, Waymo is doing pretty well on most trips (admittedly within a small geofenced zone that has friendly weather -- you can watch oodles of video from jjricks on real trips here: https://www.youtube.com/watch?v=61AfxOMu3wY https://www.youtube.com/watch?v=61AfxOMu3wY )

tempestn 5y ago
I actually have watched a number of his videos, and while it is considerably better than the Tesla self driving, it's not something I'd currently pay to experience vs driving myself (or vs an Uber at a similar price point). The driving style is still too awkward and, to me, discomfiting.
- mlyle 5y ago
  Ah, to me it looks pretty good. And there's reason to believe it'd be better on freeways. I'd definitely pay a few bucks per hour for my car to drive me to Southern California while I read a book.
  - gitanovic 5y ago
    What I don't understand is: why don't you take a train instead? I have never been to US, but whenever I hear about self driving cars I think: why don't people just use public transport if they don't want to drive?
    - gnicholas 5y ago
      Strangely, there are no direct trains from northern CA to southern CA. There is a line from Sacramento down into Santa Clara, and a line from San Luis Obispo down to San Diego. But there's no direct way to get from SF/SJ to LA/OC via train. There are buses, but they are considerably less enjoyable than the Pacific Surfliner line!
    - blhack 5y ago
      It’s too much of a hassle to use public transportation. With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc. Like for instance with my own car instead of a shared train, I can bring my golf clubs, my dog, and my drone with me, and call my friend on the way out of town to see if he’s up for a spontaneous trip to San Diego to see what fun we could get up to there for the weekend. That’s not really possible on a train/airplane. It’s just a totally different experience than arriving to a train station with some limited set of your things, possibly being late, then arriving somewhere else again without a car, and having to return again by doing all of this in reverse. It also means spending time forced to socialize with people you don’t know, and almost definitely in a seat that is less comfortable than the one in your $100k luxury car.
      - darkwater 5y ago
        > With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc. Also, you can get blocked in a traffic jam when you don't want.
tails4e 5y ago
I did think it was imminent at one point, but the fact the delivery date keeps pushing out shows its far from complete, and is only usable in controlled circumstances. I use it in traffic jams, or motorway driving, and that's it, anywhere else and it's just asking for trouble. It's useful in those limited cases, but it's a shame how many false promises have been issued, I've a lot of respect for what Elon has done, but he's really oversold this
malozite 5y ago
I think the data advantage is limited. I used to work in high-frequency equity trading. We would run simulations against tick-by-tick data from real markets. The simulations were of very questionable value. As soon as you trade once in your simulation, from that point onwards the whole run is contaminated. Your trade, if it had actually happened, would have been seen by other participants, who would have changed their strategy, often in complex and unpredictable ways, and the timelines would have diverged rapidly and irreversibly from that point. Data recorded from driving situations is useful for one thing - testing out split-second decision making - stop or accelerate? Steer left or right? This is a limited use case. Most situations don't have critical decisions like this. The ones that do, it's often hard to judge what would have been the right answer. Even accidents - where the post facto desired outcome is "don't have an accident", unfold over several seconds, involving several stages of actors responding to each others' actions. No one can be sure how the driver should have approached the very first interaction - the only one for which the data is meaningful. Actual road testing, and meaningful simulation which isn't just replaying what actually happened is much more useful.
lm28469 5y ago
> Waymo Even they admitted fully autonomous vehicles most likely won't be a thing any time soon (maybe ever if we keep our current infrastructure) https://www.cnet.com/news/alphabet-google-waymo-ceo-john-krafcik-autonomous-cars-wont-ever-be-able-to-drive-in-all-conditions/ https://www.cnet.com/news/alphabet-google-waymo-ceo-john-kra...
- mlyle 5y ago
  I feel like this was misread. The statement appears to say that he doesn't think they will completely displace manual driving of cars for considerable time. And that SAE level 5 arguably expects more than we expect from human drivers (many of us would not be safe driving in whiteout conditions on a windy mountain road).

hughrr 5y ago

So much this. I hope I don't stand on too many toes here. Actually I don't care. My father has a self driving enabled Tesla Model S which he paid a hell of a lot of money for that he doesn't have as he's in his 70s, living entirely off the remains of the house he just sold. He's convinced it'll be any day now that he gets full self driving and that it'll change the world. He tells me he loves the car and phones me daily to tell me about it. But at the end of the day there are a bunch of complaints hiding in his words. This covers everything from poor fitting, paint issues, weird bugs, massive safety and computer issues, lots of hidden costs and general battery life issues. He's hiding the true picture which is not the real experience. Between Tesla marketing and the echo chamber Tesla forums, he's traded his entire cash pile and attention for what I can only describe as a prototype that is a ball and chain which limps between bouts of charging CPR and he can't tell anyone because it might damage his conspicuous virtuous electric car identity. But Tesla will be introducing self driving any day now he tells me. All the time. For the last 3 years... I'm not really a proponent or opponent of electric cars. I don't really care about the things as they don't factor into my life past the utility of a screwdriver. I myself drive a third-hand Citroen with a tuk tuk engine in it which was cheap to buy and run and takes me to the shops and for the odd weekend out. But from my vantage point, I can't see the return of investment that such complex technology can add to a simple problem. All it does is make it more complicated, less safe and less reliable while milking people for cash for promises which may never materialise. If there's anything I've worked out over the last couple of decades in the software industry is that this model really doesn't work - it's the marketing and development model that Microsoft uses. Edit: he literally just phoned me to tell me how wonderful it is again. But it cost him 5% of the battery capacity to cool it down enough that it’s not going to burn his ass off (has sunroof) which means his next journey is going to have to end with a 10 mile detour to sit in a supercharger bay (which where he lives will be busy and he’ll have to wait) and thus probably cost a coffee at the same time.

ec109685 5y ago
Obviously you know your father best, but it at least seems from your words it makes him happy. As someone with similar car, you get to know it’s quirks and use it when it adds value.

Rygian 5y ago

What did you test-drive? Regular Tesla auto-pilot, or the limited-access FSD Beta? They are radically different experiences.

mft_ 5y ago
This is important. From the original description, my suspicion is it’s the regular (i.e. not advanced beta) “full self driving“ package, which I have on my model 3. Truth is, it’s pretty limited, and you’ve got to learn the types of situations where it will and won’t cope. Through town, situations with parked cars, or weird lane markings, it will struggle. However, the flipside is that on longer journeys – with mostly open roads, it does very well indeed. I recently did a three+ hour drive and the car drove for probably 80% of that time, and without any issues at all. Once you get used to its foibles and therefore understand when you can trust it, it really is a relaxing experience.
- Dah00n 5y ago
  The FSD beta is way worse. You have to be much more conscious about everything that happens around you than with the "old package". Then you might as well drive yourself. In my experience it's like how it feels being in a car driven by an old relative that clearly shouldn't be driving a car anymore but you're too polite to tell him.
  - ketamine__ 5y ago
    The accidents are going to increase when FSD is released to the public with subscriptions.
tempestn 5y ago
I'm not certain since it was just a test drive, and there was no Tesla rep with me - it did say "Beta" on the autopilot screen, and seemed happy to engage on any kind of street, but I don't know if that means it was the full FSD Beta. (I'd be kind of surprised if they just handed people the (virtual) keys to that with no warnings, but Tesla does tend to do surprising things.)

berkes 5y ago

I'm no expert on any of the selfdriving features (I still self-drive my 1984 Volkswagen with tlc) but from what I hear and see, Tesla is amongst the worst when it comes to selfdriving and e.g. Volvo or Volkwagen is way ahead. Their marketing, however, is top notch, so whenever Tesla fails to deliver, people conclude that 'all selfdriving must be crap'. Is there something called 'Law of inhibiting leadership'? In which an underperforming leader of a segment holds back the entire segment?

terryf 5y ago
As an owner of a Volvo V90 that has pilot assist or whatever they call it - it's great for highway driving! It has it's quirks but they are easy to get used to and after a few years, I know pretty much exactly where it is extremely reliable and where it will get it wrong. The behaviour is extremely boringly reliable - which is exactly what it should be and doing a few hundrer km on a good highway is now much much more relaxing than before.

lm28469 5y ago

Don't worry Musk said it'll be delivered in two years tops, in 2012 or something. Anyone still believing in fully autonomous cars as they are presented by Musk &co today (the real thing, not a cruise control that only works on California's highways on a sunny day) is 101% delusional. We'll talk about it in 15 years as the "previous AI hype cycle".

lvl100 5y ago

Tesla scammed people into believing in FSD. I actually believed it for a second too. Yet here we are with Elon Musk as one of the richest con of all time.