|
Anta posted:I thought Tesla was going to do some fuckery with delivery dates etc on the incentives?
|
#
?
Jul 29, 2017 23:32
|
|
|
|
| # ? Jun 4, 2024 21:05 |
|
|
Subjunctive posted:What did they tell you? From their reservation T&C: Tesla posted:1. Model 3 Reservation Basically you're getting in line for your region (fine, they disclosed that), and when your number hits you get to configure and purchase your car. They don't say that it's more like, "hey, you're getting in line for a $49k model ONLY at first and eventually, if you want the cheaper version we sold this as, then six months later you can get it."
|
|
#
?
Jul 30, 2017 01:16
|
|
|
mobby_6kl posted:I think there's also a backup keycard you can keep in the wallet or something. There is. They mentioned it in the context of a valet, but a key is a key.
|
|
#
?
Jul 30, 2017 01:26
|
|
|
CharlesM posted:I have a Windows Phone and frequently go places without cell signal. Would it still work? The new fobs and new model 3 use bluetooth low energy. most smart phones newer than 2013 or 2014 should have it. It's not an app or anything, you simple pair it like you would any bluetooth thing and when it's within range to connect, it unlocks and lets you start the car.
|
|
#
?
Jul 30, 2017 01:34
|
|
|
Roadie posted:I was interested in the whole idea of the 3 at first, but man, I'm already a lovely enough driver without having to look to the side to see the speedometer. Elon Musk was still claiming that they'd have complete self-driving capability out soon enough that no one would miss having a proper instrument panel. Which might be somewhat plausible if they can pull off the coast-to-coast autonomous trip they have planned for this year.
|
|
#
?
Jul 30, 2017 03:17
|
|
|
It's not plausible at all.
|
|
#
?
Jul 30, 2017 03:48
|
|
|
Its also kind of a lovely argument since the self-driving costs another $8000 on top of the base model.
|
|
#
?
Jul 30, 2017 05:04
|
|
|
Subjunctive posted:What did they tell you? They heavily implied, though did not outright state, that the car would be less than $35,000, and that if you got #4 in the waiting list you would get car #4, or you'd receive your car around the time the first ten people would be receiving cars. Tesla definitely did not say "you can put down your deposit but we'll bump you indefinitely unless you buy the most expensive fully-optioned version of the car." That's absolutely pulling the rug out from a lot of people, by effectively saying that they are going to move people up in line if they're willing to pay for the privilege. Why have the deposit at all, in that case? Tesla has been making misleading marketing statements that border on outright false advertising ever since Elon took over, though, so it's not particularly surprising to me that they've done it again.
|
|
#
?
Jul 30, 2017 05:41
|
|
|
Subjunctive posted:It's not plausible at all. I think it's completely plausible that they could carefully select a route that begins next to an onramp in Los Angeles, ends just off an offramp in New York, and which attempts no difficult road situations, and have the car drive autonomously coast-to-coast* spending 99% of the moving time going in a straight line on the middle lane of I-70. *not counting parking the car in Supercharger bays and/or navigating parking lots at motels and truck stops Call me when they have a car that can drive door-to-door, no human contact with the vehicle controls, from the Seattle suburbs to a New York City brownstone in the middle of January, charging itself along the way. I'll even allow a human to get out and plug in a charging cable after the car has parked itself. Then I'll call it a "fully autonomous car."
|
|
#
?
Jul 30, 2017 05:47
|
|
|
It would be interesting to see what kind of data they're actually getting from cars on the road. Still, i would bet the majority of driving done by a model 3 will be done by a human who would benefit from a basic readout they don't have to take their eyes off the road for. A slim, simple black display with speed, range and distance to turn would be nice so you could shut the screen off in the dark would be nice. I'm wondering if there's a wireless API that the aftermarket could use.
|
|
#
?
Jul 30, 2017 06:03
|
|
|
Sagebrush posted:They heavily implied, though did not outright state, that the car would be less than $35,000, and that if you got #4 in the waiting list you would get car #4, or you'd receive your car around the time the first ten people would be receiving cars. Tesla definitely did not say "you can put down your deposit but we'll bump you indefinitely unless you buy the most expensive fully-optioned version of the car." That's absolutely pulling the rug out from a lot of people, by effectively saying that they are going to move people up in line if they're willing to pay for the privilege. Why have the deposit at all, in that case? He's a shyster, and that's what makes his cult of personality so funny.
|
|
#
?
Jul 30, 2017 10:01
|
|
|
Agronox posted:From their reservation T&C: It's sort of like a combination of other manufacturers (exotics, generally) that do waiting lists, and normal manufacturers who will happily sell you what's on the lot, but if you want to custom build yours to spec, it could be months before they get that built and shipped out.
|
|
#
?
Jul 30, 2017 10:33
|
|
|
Sagebrush posted:They heavily implied, though did not outright state One of the things I just don't understand about this whole thing (and I see it a lot in comments at Ars) is this need people have to infer or project statements onto Tesla it doesn't make. And that somehow it doesn't behave like every other big company doing business. Sagebrush posted:I think it's completely plausible that they could carefully select a route that begins next to an onramp in Los Angeles, ends just off an offramp in New York, and which attempts no difficult road situations, and have the car drive autonomously coast-to-coast* spending 99% of the moving time going in a straight line on the middle lane of I-70. That's a level 5 autonomous car and anyone who tells you there will be commercially available ones of those in the next 15 years is lying or knows nothing about the topic.
|
|
#
?
Jul 30, 2017 12:04
|
|
|
Sagebrush posted:I think it's completely plausible that they could carefully select a route that begins next to an onramp in Los Angeles, ends just off an offramp in New York, and which attempts no difficult road situations, and have the car drive autonomously coast-to-coast* spending 99% of the moving time going in a straight line on the middle lane of I-70. I get your point, and the point of doing it from a technological proof of concept point of view, but Jesus why would anyone? Like, you can take a train that does that more or less today, plus an uber to get you to the door, and basically nobody does that.
|
|
#
?
Jul 30, 2017 12:51
|
|
|
drgitlin posted:That's a level 5 autonomous car and anyone who tells you there will be commercially available ones of those in the next 15 years is lying or knows nothing about the topic. Is this "level" business a Tesla marketing term or what? I've never heard it before.
|
|
#
?
Jul 30, 2017 15:39
|
|
|
TooMuchAbstraction posted:Is this "level" business a Tesla marketing term or what? I've never heard it before. SAE standard actually. https://en.wikipedia.org/wiki/Autonomous_car#Classification quote:Level Name
|
|
#
?
Jul 30, 2017 15:45
|
|
|
Sagebrush posted:and that if you got #4 in the waiting list you would get car #4, or you'd receive your car around the time the first ten people would be receiving cars. Tesla definitely did not say "you can put down your deposit but we'll bump you indefinitely unless you buy the most expensive fully-optioned version of the car." That's absolutely pulling the rug out from a lot of people, by effectively saying that they are going to move people up in line if they're willing to pay for the privilege. Why have the deposit at all, in that case? I'm about 95% sure that they were clear from the beginning that higher specced cars would be built first and that the actual $35,000 models would be coming later. That and the dashboard were the two main reasons I decided to take the money I was going to put down for one and buy a Vive instead. If Tesla themselves didn't explicitly say that then it was still so widely assumed in the community as to be basically considered fact. Didn't they also do the same thing for the S and X launches? The first production models were drat near loaded, weren't they?
|
|
#
?
Jul 30, 2017 16:20
|
|
|
wolrah posted:
Yep.
|
|
#
?
Jul 30, 2017 16:24
|
|
|
drgitlin posted:That's a level 5 autonomous car and anyone who tells you there will be commercially available ones of those in the next 15 years is lying or knows nothing about the topic. There's some confusion because the NHTSA and SAE came up with two different definitions but both used "level". NHTSA level 4 is SAE level 5. There is no NHTSA level 5. To expand on what Ola posted -- (SAE definitions) Level 0: No automation, you do everything yourself. Level 1: A single channel of automation that allows the car to perform either steering or acceleration/deceleration on its own, with the human performing the rest of the driving task. Functionally this means a car with speed-holding or range-based cruise control. Level 2: Multiple channels of automation that allow the car to perform both steering and acceleration/deceleration, with the human performing the rest of the task. Today, this means a car with lane-following steering assist, distance-keeping cruise control, and automatic braking. Teslas have this as "autopilot", and every other manufacturer has it as well as "driver assists" in their high-end cars. The car is not expected to do anything except keep itself in the lane and accelerate/decelerate appropriately, so the human needs to be constantly ready to take control at a moment's notice (i.e. keep your hands on the wheel). Emergencies are the driver's responsibility. Level 3: The same as level 2, but the car can perform the driving tasks for "extended periods." The car can make a request for driver intervention if the situation goes beyond its capacity, and the driver is expected to take over. The key difficulty with level 3 is that, because the car is driving itself for half an hour at a time or more, it's anticipated that the driver will zone out. The car therefore has to be able to safely handle the transition between autonomous and human control over a few seconds, which is an eternity in an emergency. Most manufacturers are aiming to skip over level 3 because of anticipated legal problems with the transition of control authority. Tesla claims that they are at level 3 right now because you can drive on autopilot on the highway for a long time, but by saying that people must always keep their hands on the wheel, they're actually at level 2. Level 4: The car can perform the driving task for extended periods and correctly handle all "dynamic driving situations," even if the driver does not respond to a request for intervention. Basically this means fully hands-off driving, including reliable handling of emergencies, maintaining lane position through inference when the lane markings are worn off, following flagmen at construction sites, avoiding dangerous potholes, and the like. It is okay to limit the envelope in which the car can do this, so you can say e.g. "if there's a whiteout snowstorm, the car pulls itself over." Obviously, assigning any major limitation on the envelope such as "the car can only drive on roads where the lane markings are bright and clean" means it's pretty weasely to call it a self-driving car, though that is technically allowed. Level 5: The car can do anything a human driver can do in any environmental condition that a human can handle, such as parking the car correctly beside others in a gravel lot, or driving on an unmarked dirt road. Requires hard AI. Elon Musk has stated that Tesla will have level 4 cars ("don't touch the controls from coast to coast, parking lot to parking lot") by the end of the year, and level 5 ("fall asleep in the car and wake up at your destination") two years from then. edna_krabappel_ha.wav
|
|
#
?
Jul 30, 2017 18:38
|
|
|
No one, even NHTSA, still uses the NHTSA scale.
|
|
#
?
Jul 30, 2017 18:48
|
|
|
Sagebrush posted:Level 5: The car can do anything a human driver can do in any environmental condition that a human can handle, such as parking the car correctly beside others in a gravel lot, or driving on an unmarked dirt road. Requires hard AI. "Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't drive at least as well as a human in all situations a human would encounter?
|
|
#
?
Jul 30, 2017 18:54
|
|
|
TooMuchAbstraction posted:"Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't drive at least as well as a human in all situations a human would encounter? It needs to able to problem solve and cooperate safely with humans in situations that the programmers haven't taught it, that the magic machine learning hasn't learned yet, better than humans do, measured in accident statistics. That's pretty drat close to full AI, even if it's limited to just driving. I'm skeptical. I think it's misguided. Why is it so drat important to send an empty car from A to B? What will traffic be like when every car trip contains both a manned and an unmanned leg? Isn't that a massive waste of energy, even if they're all electric? It feels like 50s-60s futurism, we've watched an episode of the The Jetsons and thought, yes one day daddy will commute by spaceship to the office on the moon and homemaking mommy will get an atomic vacuum cleaner for space Christmas, it seems so inevitable. I think we're way way off the mark on this one. We suck at predicting our future at the best of times, and we have always overestimated the computer as a machine of independent intelligence and always underestimated what new paradigms of human interaction suddenly appear out of nowhere instead.
|
|
#
?
Jul 30, 2017 19:10
|
|
|
TooMuchAbstraction posted:"Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't cook a hamburger at least as well as a human in all situations a human would encounter? TooMuchAbstraction posted:"Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't handle customer service at least as well as a human in all situations a human would encounter? TooMuchAbstraction posted:"Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't perform surgery at least as well as a human in all situations a human would encounter? TooMuchAbstraction posted:"Hard AI" is a pretty strong claim. Is there some reason you don't think a sufficiently advanced expert system couldn't write music at least as well as a human in all situations a human would encounter? This is why "sufficiently advanced" is a meaningless metric. We've never developed an "expert system" that can do anything "in all situations a human would encounter" as well as a human would. It's Zeno's paradox made real -- the closer we get, the more we discover that there's still more distance to cover. Sure, eventually we might invent something that can do it. I believe that system will be a lot closer to a true human intelligence than an advanced computer. Look at something as apparently-simple as a conversational "AI" for an example. ELIZA was invented in the 1960s and everyone's minds were blown. All she did was repeat your own phrases back to you "I'm eating a banana right now." "Interesting. What do you think of bananas?" but it initially fooled everyone who saw it and, for a brief time, passed the Turing test. Then people quickly caught on to what was happening and whoops, now the only people ELIZA can fool are 5-year-olds with no understanding of the patterns. Every subsequent conversational AI went through the same cycle, and continues to do so today. Self-driving cars are going to get better and better, and we're going to discover more and more situations in which they don't work and more development is still required, and on and on. Sagebrush fucked around with this message at 19:20 on Jul 30, 2017 |
|
#
?
Jul 30, 2017 19:18
|
|
|
Ola posted:It needs to able to problem solve and cooperate safely with humans in situations that the programmers haven't taught it, that the magic machine learning hasn't learned yet, better than humans do, measured in accident statistics. That's pretty drat close to full AI, even if it's limited to just driving. quote:I'm skeptical. I think it's misguided. Why is it so drat important to send an empty car from A to B? What will traffic be like when every car trip contains both a manned and an unmanned leg? Isn't that a massive waste of energy, even if they're all electric? I don't think it's misguided, but I also don't think that we'll be sending empty cars all over the place all the time. I mean, sure, you'd have the car drop you off and then go park itself, and that would have some impact on traffic, but it'd be massively overshadowed by the other things that driverless cars enable. In particular, granting vastly improved mobility to the millions of elderly, drunk, disabled, etc. people who are not able to safely control a vehicle on their own. I don't know what the percentages are on that, but even adding 5% more traffic to the roads from people driving that wouldn't normally have been able to is going to have a substantial impact on traffic, which is usually (at least where I live) constantly teetering on the edge of being oversubscribed. As for wastes of energy, sure, in a purely kinetic sense it is. On the other hand, it's a pretty big waste of energy to demand that a person devote their entire attention to the task of getting from point A to point B, when there's other things they could be doing with their time. Imagine how much happier your average overworked office drone would be if they could nap in the car during their commute. ...of course, being overworked office drones, they'd probably spend the time working instead. 
|
|
#
?
Jul 30, 2017 19:23
|
|
|
TooMuchAbstraction posted:If what you describe were actually the case, then I'd agree with you. But I don't think that a car-driving program has to be able to cope with literally innumerable kinds of situations to be able to thoroughly trounce humans in terms of accident statistics. Well, the situations you can encounter in traffic today are literally innumerable, you cannot put a number on the amount of traffic situations you could possible encounter because you can always add or tweak one of the factors. But you as a human can sense and understand what's going on and make some sort of decision to get you through it - even if you only partly understand it and you can even cancel the trip. One thing I love using cars for is going to the mountains in winter to go skiing. Autopilot or adaptive cruise is notoriously susceptible to interference by snow. If I used a self driving car service to go to and from the mountain, it would have to be as good as me to see through the snow, keep its sensors clean, avoid snow drifts while not hitting oncoming traffic, avoid spinning the tires here but use spinning to your advantage there and so on. Even then, it won't be able to go out and dig away snow if the car gets stuck. That's not too much of an edge case, the mountains are full of people every winter who easily, even the quite retarded ones, utterly trounce anything a computer ever has come close to when it comes to navigating a car through challenging conditions. And no Silicon Valley genius is bragging about the awesome snow drift avoidance algorithm they've made, those real world scenarios are too boring and un-Jetsons to even consider. That's why I think level 2 is great for comfort, level 4 is possible for particular highways, but level 5 is a Jetsons fantasy and not even that desirable - it may at once be safer than human drivers and cause more deaths, because it will be covering more distance and thus exposed to more risk.
|
|
#
?
Jul 30, 2017 19:52
|
|
|
I went to look up some stuff about the Tesla autopilot since I'm not sure exactly how far its capabilities currently go re. stuff like handling right-of-way at four-way stops or reading temporary signage, both of which are mundane daily driving tasks which cannot be described with a simple if-this-then-that algorithm. The Tesla forums are a reminder that, regardless of what Tesla actually says and what is realistically possible, the greater public has next to no understanding of how these systems work:quote:johndoeeyed | October 22, 2016 quote:burnrubber | October 20, 2016 quote:Bighorn | October 20, 2016
|
|
#
?
Jul 30, 2017 22:00
|
|
|
Ola posted:Well, the situations you can encounter in traffic today are literally innumerable, you cannot put a number on the amount of traffic situations you could possible encounter because you can always add or tweak one of the factors. But you as a human can sense and understand what's going on and make some sort of decision to get you through it - even if you only partly understand it and you can even cancel the trip. Sure, but "literally innumerable" doesn't mean the same thing as "impossible for a computer to cope with", or else computers wouldn't be able to e.g. calculate a ballistic trajectory, something they've been doing since before they were electronic. The question is if there are qualitatively innumerable situations, ones which require novel thinking that a human could come up with but a computer can't. And I'm not convinced there are. There are plenty of very difficult scenarios, like your example of driving on a mountain in a blizzard. And I won't for an instant claim that this is easy or achievable in any kind of near-term timespan. I'm just saying I'm not convinced that we have to solve the hard AI problem in order to produce a computer that can drive as well as a skilled human. Sagebrush posted:The Tesla forums are a reminder that, regardless of what Tesla actually says and what is realistically possible, the greater public has next to no understanding of how these systems work. You think that's bad? The software developers designing these systems don't really understand how they work, either. 
|
|
#
?
Jul 30, 2017 22:11
|
|
|
TooMuchAbstraction posted:If what you describe were actually the case, then I'd agree with you. But I don't think that a car-driving program has to be able to cope with literally innumerable kinds of situations to be able to thoroughly trounce humans in terms of accident statistics.
|
|
#
?
Jul 30, 2017 22:59
|
|
|
TooMuchAbstraction posted:The question is if there are qualitatively innumerable situations, ones which require novel thinking that a human could come up with but a computer can't. One of us might be right, or not. We'll see. One way to solve this is to limit the area (and perhaps time/season) where self driving cars are allowed and support them with public infrastructure in addition to the car's own sensors. Then you have level 4, and people should be happy with that. It's not a car that can drive itself through the Amazon forest, but you can go to sleep in it and if the destination is outside the autonomous network, it pulls up in a parking lot like a plane on autopilot entered a holding pattern.
|
|
#
?
Jul 30, 2017 23:00
|
|
|
ilkhan posted:To be fair, humans are really bad at unexpected situations, and moderately bad at routine situations. They don't have to be perfect to meaningfully trounce humans. I disagree with this. We might be bad at reacting quickly and precisely, but if we have time on our side we are extremely good at improvising, problem solving and finding solutions - compared to animals and a computer car encountering a moose or a flash flood.
|
|
#
?
Jul 30, 2017 23:06
|
|
|
It's also worth noting that there are a lot of human drivers who at best achieve Level 4 poorly. I know a lot of people who will just pull over and stop when things get hard and are almost certainly less predictable than a computer in the easier situations.
|
|
#
?
Jul 30, 2017 23:28
|
|
|
I love this topic, but it's always contentious and there's a thread for it somewhere (no search on mobile) which is better suited for continuing the conversation.
|
|
#
?
Jul 30, 2017 23:31
|
|
|
TooMuchAbstraction posted:Sure, but "literally innumerable" doesn't mean the same thing as "impossible for a computer to cope with", or else computers wouldn't be able to e.g. calculate a ballistic trajectory, something they've been doing since before they were electronic. The question is if there are qualitatively innumerable situations, ones which require novel thinking that a human could come up with but a computer can't. And I'm not convinced there are. There are plenty of very difficult scenarios, like your example of driving on a mountain in a blizzard. And I won't for an instant claim that this is easy or achievable in any kind of near-term timespan. I'm just saying I'm not convinced that we have to solve the hard AI problem in order to produce a computer that can drive as well as a skilled human. To elaborate on your example of calculating a ballistic trajectory -- that is something humans are capable of as well. And not just by churning through calculus; any time you throw a ball, you're making calculations of mass, force, angle, wind, and so on, using a neural network. With a few years of practice, any human child can learn to spontaneously throw a ball and hit a moving target. It took the entire human race fifty thousand years to develop a technology capable of replicating that ability, and only in the last few decades do we finally have machines that can do it faster and better than a human. One projectile, one target, no confounding factors. Obviously this is why neural networks seem to have so much promise, but we've yet to develop anything that even resembles a "human driving a car" neural model (which is what I am referring to when I say "hard AI"; obviously the car AI doesn't have to be able to write music or do standup comedy or whatever). The neural networks that we see demonstrated are discrete systems that handle things like "find all the signs and read the numbers", all feeding their data into a central driving model that is relatively fixed in its behavior. It does a pretty good job of replicating a human's ability to collect data from a variety of sources and produce driving inputs, but it's still inherently limited by its training set. I've yet to see anything that approaches the human capacity to synthesize new (generally correct) rules on the fly, where people take specific situational experience, generalize a rule, and reapply it on the specific level in a fraction of a second (this is something we can see happening in real-time with fMRI). And, of course, computerized neural networks still all suffer from the issues mentioned above like blackboxing and training bias. Imagine if you had raised a kid completely in isolation in a black room with a picture of a target moving back and forth on one wall, and then when he'd learned how to hit that target, you took him out to play baseball. That's pretty much how a current-gen neural network is trained, and it doesn't really generalize well for obvious reasons.
|
|
#
?
Jul 30, 2017 23:44
|
|
|
To be on topic, looks like Nissan will be giving the rest of the 2018 Leaf info out on Sept 5th. Meanwhile, they've added a mode to make the accelerator be an analog controller for the speed.
|
|
#
?
Jul 30, 2017 23:46
|
|
|
Subjunctive posted:I love this topic, but it's always contentious and there's a thread for it somewhere (no search on mobile) which is better suited for continuing the conversation. If this thread actually exists, I'd be interested in a link as I can't find it in search  (I work in the industry and always like to be a fly on the wall in this kind of talk)
|
|
#
?
Jul 31, 2017 01:08
|
|
|
I don't follow this thread in general. Has it been beaten to death how people are defending the Model 3 not having a speedometer on the dash by saying it would make the car cost more? Because JFC that's so loving stupid. You can argue the necessity of it, but it couldn't cost more than another $10 per unit.
|
|
#
?
Jul 31, 2017 05:51
|
|
|
Sagebrush posted:Level 2: Multiple channels of automation that allow the car to perform both steering and acceleration/deceleration, with the human performing the rest of the task. Today, this means a car with lane-following steering assist, distance-keeping cruise control, and automatic braking. Teslas have this as "autopilot", and every other manufacturer has it as well as "driver assists" in their high-end cars. The car is not expected to do anything except keep itself in the lane and accelerate/decelerate appropriately, so the human needs to be constantly ready to take control at a moment's notice (i.e. keep your hands on the wheel). Emergencies are the driver's responsibility. I thought Tesla isn't supposed to call it autopilot anymore after it killed somebody in China and the family sued. 22 Eargesplitten posted:I don't follow this thread in general. Has it been beaten to death how people are defending the Model 3 not having a speedometer on the dash by saying it would make the car cost more? Because JFC that's so loving stupid. You can argue the necessity of it, but it couldn't cost more than another $10 per unit. It'd probably cost significantly more than $10 per unit. Tesla is pinching pennies like crazy because they're almost certainly losing money on the initial model 3 units.
|
|
#
?
Jul 31, 2017 06:29
|
|
|
Gamesguy posted:I thought Tesla isn't supposed to call it autopilot anymore after it killed somebody in China and the family sued. Can you explain how they might lose money on initial units then make a profit on units later on? Is it purely economies of scale, variable costs becoming cheaper?
|
|
#
?
Jul 31, 2017 07:01
|
|
|
22 Eargesplitten posted:Has it been beaten to death how people are defending the Model 3 not having a speedometer on the dash by saying it would make the car cost more? No. Gamesguy posted:I thought Tesla isn't supposed to call it autopilot anymore after it killed somebody in China and the family sued. No.
|
|
#
?
Jul 31, 2017 07:07
|
|
|
|
| # ? Jun 4, 2024 21:05 |
|
|
22 Eargesplitten posted:I don't follow this thread in general. Has it been beaten to death how people are defending the Model 3 not having a speedometer on the dash by saying it would make the car cost more? Because JFC that's so loving stupid. You can argue the necessity of it, but it couldn't cost more than another $10 per unit. I predict a thriving market in GPS based stick-on speedometers that match the model 3's dash texture/color. Kind of like the A pillar gauge pods that were big in the 90's
|
|
#
?
Jul 31, 2017 15:21
|
|