# Self driving cars are still decades away **URL:** \ **Category:** In My Humble Opinion\ **Created:** [October 10, 2017, 5:38pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499 "2017-10-10T17:38:25Z")\ **Posts on this page:** 20\ **Page:** 99
**Author:** ![TroutMan](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/troutman/32/6721_2.png) [@TroutMan](https://boards.straightdope.com/u/TroutMan)\ **Post date:** [June 20, 2024, 5:13pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1968 "2024-06-20T17:13:10Z")
> [@hajario](#): > > During FSD the driver needs their hand on the wheel and to be facing forward. Maybe a blowjob but that’s about it. There are other positions beyond a blowjob that allow the driver to face forward with one hand on the wheel. It’s a popular genre on homemade porn sites. ---
**Author:** ![hajario](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hajario/32/171_2.png) [@hajario](https://boards.straightdope.com/u/hajario)\ **Post date:** [June 20, 2024, 5:21pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1969 "2024-06-20T17:21:21Z")
I’m going to take you at your word on this ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [June 20, 2024, 6:04pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1970 "2024-06-20T18:04:28Z")
> [@TroutMan](#): > > There are other positions beyond a blowjob that allow the driver to face forward with one hand on the wheel. It’s a popular genre on homemade porn sites. Rule 442, there’s even porn for sports cars! ---
**Author:** ![steronz](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/steronz/32/11765_2.png) [@steronz](https://boards.straightdope.com/u/steronz)\ **Post date:** [June 20, 2024, 6:10pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1971 "2024-06-20T18:10:33Z")
I’m guessing they’re remembering things wrong. [Sixt recently dumped Tesla](https://jalopnik.com/rental-company-sixt-will-begin-dumping-tesla-fleet-due-1851081220), following Hertz’s lead (albeit more aggressively), but it was due to maintenance costs and, probably implied, higher than expected depreciation. Anyone who bought a bunch of Teslas prior to the price cuts and resultant affect on the used car market is probably scrambling. ---
**Author:** ![Gozu](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/gozu/32/6615_2.png) [@Gozu](https://boards.straightdope.com/u/Gozu)\ **Post date:** [June 21, 2024, 3:04am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1972 "2024-06-21T03:04:00Z")
[quote=“echoreply, post:1959, topic:798499, full:true”] > [@TroutMan](#): > > It’s that a system like FSD that _almost_ provides autonomous driving except in rare situations causes people to lose focus. They are physically not able to react to something as quickly as if they were actively driving People are unable to pay attention when they are actively driving. All of the arguments about vigilance tasks also apply (to some degree) to actively driving. > [@hajario](#): > > Per mile, which system causes fewer injuries/deaths/accidents? That’s all that matters. cars are always going to kill people. Which is the safer system? > > This is the real question, and currently we don’t have data to answer it. Tesla probably has data to answer it, but they really don’t release enough to trust the answer. Gee. I wonder why Tesla doesn’t release that data! 😋 Let’s try and figure it out using logic: If the data showed that FSD drives better than a human, as calculated by accidents, injuries and deaths per million miles, would Tesla release it and use it as a selling point? Yes. Confidence level: 99.999% Corollary: Does that mean that the actual numbers would make Tesla FSD look bad? Yes. That’s it. That’s the end of my logic. Waymo did release data that showed their cars were approximately 4x safer than humans. Why? Because in America if you got it you flaunt it. Corporations are not humble. I’m a Tesla owner and I thought it was both offensive and laughable for Tesla to ask ME to pay them $6k to train their AI, putting my life at risk in the process. Not even beta test it mind you! Straight up train… I tried it when they offered it for free for a month and it is dangerous because it works well 99% of the time. That’s enough to lull you to distraction but bad enough to get you killed. Something that has happened multiple times already. I repeat: it is dangerous. Human brains have certain flaws incompatible with FSD as it is. It’s either lane assist + adaptive cruise or L4 for us. Nothing in between is safe. Google/waymo realized this a long time ago and they were unfortunately correct. ---
**Author:** ![Dr.Strangelove](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dr.strangelove/32/6613_2.png) [@Dr.Strangelove](https://boards.straightdope.com/u/Dr.Strangelove)\ **Post date:** [June 21, 2024, 3:09am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1973 "2024-06-21T03:09:04Z")
> [@Al128](#): > > so … I am not seeing this happen on a wider scale … Turo already exists for non-self-driving cars and I don’t get the impression that trashed cars are a huge problem. I’m sure it happens sometimes, but users get ratings tied to their ID/credit card, and a deposit system can handle cleaning/repairs if required. There are plenty of ways to handle this kind of thing. Also, the cars have interior cameras. > [@Folacin](#): > > If Tesla tells me that they have achieved L5 (for whatever definition), then I’d say they are warrantying that the system will handle whatever is thrown at it (including things potentially actually being thrown at the car). Well, I don’t know exactly what Ed was suggesting. I took it to mean basically anything that could happen to a car while the driver isn’t inside. Tesla might take responsibility for things directly related to the functionality of the system, like if a car blows through a stop light and causes an accident. But if someone throws a cinder block from an overpass, and the car obviously can’t do anything? No way will Tesla (or any other manufacturer) take responsibility for that no matter what level they’re advertising. Only if it’s part of the robotaxi fleet, where their insurance would take over. There’s some gray area here and it’ll probably take the courts to hash it out. They may end up having to compare to what a reasonable person could have done. If the car slams into another car in broad daylight, that’s obviously a problem. What if it happens at night and the other car didn’t have its headlights on? It’s much more questionable then. ---
**Author:** ![Al128](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/al128/32/16541_2.png) [@Al128](https://boards.straightdope.com/u/Al128)\ **Post date:** [June 21, 2024, 3:56am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1974 "2024-06-21T03:56:15Z")
> [@Maserschmidt](#): > > Can you provide a link to a news article about that Austria story? I only did a casual google, but did not find it. > **[Eloop beendet Tesla-Carsharing und wird Insolvenz beantragen](https://www.derstandard.at/consent/tcf/story/3000000217899/eloop-beendet-tesla-carsharing-und-wird-insolvenz-beantragen)** > > Die Flotte aus rund 200 Teslas wird aufgelöst. Eloop-CEO Leroy Hofer macht auch das Verhalten mancher User für den Schritt verantwortlich: die Fahrzeuge wurden teils mit schweren Schäden retourniert ---
**Author:** ![Al128](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/al128/32/16541_2.png) [@Al128](https://boards.straightdope.com/u/Al128)\ **Post date:** [June 21, 2024, 4:12am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1975 "2024-06-21T04:12:49Z")
> [@hajario](#): > > People trash rental cars. They always have. How is this a Tesla specific thing? it seems that the crazy fast 0-100km/h times attracted the fast-and-furious crowd, who can inexpensively experience high-performance characteristics for the price of a Nissan-Versa-rental. Donuts galore, street racing, …etc… The operation was a Rental / Car Sharing hybrid, where they had a hard time to find out who used 10,000 miles equivalent of tires in a few hours, etc… > [@leahcim](#): > > I could see Teslas being something of a status symbol and renting one would lead to increased sexy-time in a way that a Toyota Corolla wouldn’t. that! they were used as “chick-magnet” on the weekends in the (low end) clubbing scene - not necesarily sex while you drive but more like “party hard in a car that is not mine”, then do a couple of 1/4 mile runs to impress the ladies, hold my RedBull and watch this … IOW: live the fast-and-furious lifestyle for $50.00 > [@Dr.Strangelove](#): > > Turo already exists for non-self-driving cars and I don’t get the impression that trashed cars are a huge problem. I’m sure it happens sometimes, but users get ratings tied to their ID/credit card, and a deposit system can handle cleaning/repairs if required. There are plenty of ways to handle this kind of thing. Also, the cars have interior cameras. according to the news-article (and some 500 posts beneath it) I posted, this was a problem … lots of expensive damage by inexperienced drivers that were not the ones that rented the car (car handed over to a friend), and the rental/car sharing company would have to demand basically 20 year olds to recoup $30k worth of damages, while having to pay out of pocket for extremely expensive repairs. Also IIRC it was mentioned that repair-turnaround was slow and cars were out of commision for a longer time. ---
**Author:** ![puzzlegal](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/puzzlegal/32/3827_2.png) [@puzzlegal](https://boards.straightdope.com/u/puzzlegal)\ **Post date:** [June 21, 2024, 4:41am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1976 "2024-06-21T04:41:26Z")
I’m trying to catch up. I just returned from a trip where i visited a friend who was an early Tesla adopter and a huge Tesla fanboy. He has full self driving, and his car drove me all around the greater Chicago area. It was… an interesting experience. In the one hand, i was impressed with what the car can do. In the other, it’s clearly not ready for primetime yet. ---
**Author:** ![hajario](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hajario/32/171_2.png) [@hajario](https://boards.straightdope.com/u/hajario)\ **Post date:** [June 21, 2024, 5:15am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1977 "2024-06-21T05:15:00Z")
> **[FSD Is 5X Safer than Driving Old Line Teslas & 18X Safer than the Average...](https://www.tesmanian.com/blogs/tesmanian-blog/fsd-is-5x-safer-than-driving-old-line-teslas-and-18x-safer-than-the-average-car-in-the-us-underline-ark)** > > After analyzing Tesla's accident data, the analyst in Ark Invest concluded that Full Self-Driving is 5X Safer than driving old line Teslas and 18X safer than the average car in the US. I can’t speak to its accuracy but there is data out there. Here is data published by Tesla. How about that. I am even less sure about its accuracy but to say that they don’t publish data is wrong. > **[Tesla publishes astonishing FSD safety statistics](https://www.teslarati.com/tesla-fsd-safety-statistics/)** > > Tesla has published a new report detailing just how safe Full Self Driving (FSD) is making the driving experience of its customers. BTW, you screwed up your coding and it wasn’t really me that you quoted. ---
**Author:** ![hajario](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hajario/32/171_2.png) [@hajario](https://boards.straightdope.com/u/hajario)\ **Post date:** [June 21, 2024, 5:18pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1978 "2024-06-21T17:18:05Z")
I’m glad that you got to experience it. As you know, it can only get better by collecting data from a large amount of users. ---
**Author:** ![Gozu](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/gozu/32/6615_2.png) [@Gozu](https://boards.straightdope.com/u/Gozu)\ **Post date:** [June 22, 2024, 5:01am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1979 "2024-06-22T05:01:19Z")
You are right to not blindly trust its accuracy. It is very much in question. > **[Tesla Again Paints A Crash Data Story That Misleads Many Readers](https://www.forbes.com/sites/bradtempleton/2023/04/26/tesla-again-paints-a-very-misleading-story-with-their-crash-data/)** > > Tesla's crash data seem amazing, but the reality is very different and many have been misled by the numbers. In actuality, there's no strong case that Teslas are much safer with Autopilot on > Tesla’s number give a very incorrect impression — so incorrect that it is baffling why they publish them when this has been pointed out many times by many writers and researchers. Oddly, Tesla has the real data — they have the best data in the world about what happens to their vehicles. The fact that they could publish the truth but decline to, and instead publish numbers which get widely misinterpreted raises the question of why they are not revealing the full truth, and what it is that they don’t reveal. I must agree, and ask: What are the detailed disengagements statistics? How many times per drive does FSD drive horribly? I KNOW it’s not zero. In fact I know it was several times per drive for me. It tried to drive me at 60 MPH through an exit toll that was limited to 25! It couldn’t pick a gosh dang lane when turning left with 2 turning lanes! it frigging straddled them and caused people to honk at me before I disengaged. That is ridiculous. Straddling 2 lanes? Seriously? For several seconds for no reason? Embarrassing… It cut people off on the highway for no reason. it switched lanes when it was stupid to do so. The list was fucking endless. My SO kept blaming me for dangerous crap the FSD did because I didn’t react instantly to override it. ps: Yeah, the coding is tricky and I made the embarrassing mistake to post on mobile. 😅 ---
**Author:** ![Dr.Strangelove](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dr.strangelove/32/6613_2.png) [@Dr.Strangelove](https://boards.straightdope.com/u/Dr.Strangelove)\ **Post date:** [June 22, 2024, 5:28am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1980 "2024-06-22T05:28:13Z")
> [@Gozu](#): > > I must agree, and ask: What are the detailed disengagements statistics? How many times per drive does FSD drive horribly? I KNOW it’s not zero. The short and obvious answer is that they don’t publish it because it would make them look bad. But the more nuanced position is that the numbers wouldn’t be very useful, and would paint them in an _artificially_ negative light. And that Forbes blogspammers would latch on and put their own negative spin on things. We don’t have that figure for Waymo either, and _can’t_, because they don’t have drivers. We know that number isn’t zero either, or anywhere close to their published disengagement rate, because there are stories galore of the cars behaving in annoying ways that a human wouldn’t, but eventually figuring themselves out. Like the cars getting stuck at a green light. If it were FSD, a human would have intervened 100% of the time. With the Waymo… probably counts as nothing as long as the car eventually starts moving. And maybe not even when there’s human intervention. In short, Tesla has no motivation to publish these numbers because they can’t be realistically compared to anything else. Hell, it can’t even be compared to a _human_, because on any given drive with someone else at the wheel, there are going to be several times where they do something I wouldn’t, and probably would have intervened had FSD done the same thing. FSD isn’t as good as Waymo overall, but even if they were, the intervention rate would look terrible, because people are likely to take over for relatively minor issues that have no serious safety impact. And that’s something that Waymo is probably incapable of measuring even in principle, because if they could detect those events reliably, they could probably fix them as well. ---
**Author:** ![hajario](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hajario/32/171_2.png) [@hajario](https://boards.straightdope.com/u/hajario)\ **Post date:** [June 22, 2024, 5:59am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1981 "2024-06-22T05:59:23Z")
I’ll take you at your word as far as your experience but I have over 1000 miles of FSD driving and it’s done something weird maybe six times. It’s not even close to the Mr. Toad’s nightmare that you experienced. ---
**Author:** ![wguy123](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/wguy123/32/3161_2.png) [@wguy123](https://boards.straightdope.com/u/wguy123)\ **Post date:** [June 22, 2024, 1:27pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1982 "2024-06-22T13:27:08Z")
> [@Dr.Strangelove](#): > > Hell, it can’t even be compared to a _human_, because on any given drive with someone else at the wheel, there are going to be several times where they do something I wouldn’t, and probably would have intervened had FSD done the same thing. This is a really good point. Don’t tell my wife, but if she was FSD and I could intervene, I would… A LOT! And the reverse is true: She doesn’t like how I drive at times. I’m much more aggressive and she is way too timid. ---
**Author:** ![puzzlegal](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/puzzlegal/32/3827_2.png) [@puzzlegal](https://boards.straightdope.com/u/puzzlegal)\ **Post date:** [June 22, 2024, 2:05pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1983 "2024-06-22T14:05:26Z")
> [@hajario](#): > > I’ll take you at your word as far as your experience but I have over 1000 miles of FSD driving and it’s done something weird maybe six times. It’s not even close to the Mr. Toad’s nightmare that you experienced. Huh. I spent a weekend with a friend and it did 3 really weird things, and several others that he overrode: We went through a tollbooth, the speed limit went back to “highway”, and the car puttered along at 15mph, causing other drivers to swerve dangerously around it. Josh should have overridden it well before he did. It saw a ball roll onto the road 100 yards away. It stopped. Then it started moving again. As we got close, with the child watching in horror, the car would have driven over the ball (probably damaging it) if Josh hadn’t stopped it. It played “you go, nope, I’m moving” with a woman at an intersection about 6 times before Josh overrode it and went through the intersection. Josh said the car started creeping forward because the woman looked away. Might be, i thought she was glancing at the shift lever as she was about to go. I suppose that one wasn’t actually dangerous, but it confused the hell out of the other driver and must have been annoying as hell for her. I also saw it go too fast, too slow, and cut off other drivers. It’s amazing what it can do, but it’s not ready for prime time. ---
**Author:** ![Gozu](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/gozu/32/6615_2.png) [@Gozu](https://boards.straightdope.com/u/Gozu)\ **Post date:** [June 23, 2024, 3:24am UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1984 "2024-06-23T03:24:07Z")
> [@Dr.Strangelove](#): > > > [@Gozu](#): > > > > I must agree, and ask: What are the detailed disengagements statistics? How many times per drive does FSD drive horribly? I KNOW it’s not zero. > > The short and obvious answer is that they don’t publish it because it would make them look bad. > > But the more nuanced position is that the numbers wouldn’t be very useful I disagree about the numbers not being useful. At the very least, if we get consistent numbers over time, we can see if things are improving or not. And if they are improving, how much? When Elon says there is 5x-10x fewer disengagement for the latest version, that was exactly what he was doing, comparing numbers over time. …Except we can’t tell if he’s lying or not, and of course he’s probably lying because he has much less credibility than the boy who cried wolf at this point. After all, the boy only did it twice before he was eaten 😉 And so I’m saying to Musk: Put up or shut up, I am several years past my breaking point with his confident “predictions”. I actually do not have any love for hubris and lies, even if one calls them pathological optimism or whatever other misleading euphemism they come up with. As far as Tesla vs Waymo, I am pretty sure Waymo cars haven’t killed a single person **yet**. The same is not true of Tesla. Self-driving cars solely exist to eliminate deaths. Any comparison of the two tech stacks makes it obvious Waymo is winning and there is no indication this will change anytime in the foreseeable future. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [June 23, 2024, 12:40pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1985 "2024-06-23T12:40:01Z")
> [@Gozu](#): > > When Elon says there is 5x-10x fewer disengagement for the latest version, that was exactly what he was doing, comparing numbers over time. Apologies for the complete tangent, but when someone says 5x or 10x fewer, do they mean one fifth or one tenth as many? Is that a common usage? I know Gozu is just quoting Musk, throwing this out to anybody. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [June 23, 2024, 1:02pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1986 "2024-06-23T13:02:13Z")
> [@Maserschmidt](#): > > Apologies for the complete tangent, but when someone says 5x or 10x fewer, do they mean one fifth or one tenth as many? Is that a common usage? Yes, that construction has become so, much to my mild chagrin. But it’s such common usage now (and maybe was for a long time), so might as well get used to it. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [June 23, 2024, 1:09pm UTC](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499/1987 "2024-06-23T13:09:18Z")
> [@pulykamell](#): > > Yes, that construction has become so, much to my mild chagrin. But it’s such common usage now (and maybe was for a long time), so might as well get used to it Get used to it? I’ll embrace it! I’ll complain, um, 10x or 20x less frequently about it. (though I’ve never actually complained about it, so the math there gets complicated) Anyway, thanks, and back to our regularly-scheduled programming. [Previous page](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499.md?page=98) [Next page](https://boards.straightdope.com/t/self-driving-cars-are-still-decades-away/798499.md?page=100)