Hacker News new | past | comments | ask | show | jobs | submit login
Whistleblower drops 100 GB of Tesla secrets to German news site (jalopnik.com)
586 points by VagueMag 4 days ago | hide | past | favorite | 594 comments

Is it just me, or are these numbers almost low enough to be labelled positive PR?

> The Tesla files contain more than 2,400 self-acceleration complaints and more than 1,500 braking function problems, including 139 cases of unintentional emergency braking and 383 reported phantom stops resulting from false collision warnings. The number of crashes is more than 1000. A table of incidents involving driver assistance systems where customers have expressed safety concerns has more than 3000 entries.

Out of millions of cars? And for the entire time Tesla’s been making cars?

I’ve watched YouTube videos of a SF Tesla driver who submits one or two reports every video (autopilot wasn’t aggressive enough, or failed to navigate a complicated turn, etc). Are each of those counted in the 3000 safety concerns?

I can’t even fathom any other car manufacturer having lower numbers.

People are just not reporting them. My model 3 does something scary at least a few times a month but I've never reported any.

Number of actual accidents per million miles is the important stat though, and Tesla appears to be doing reasonably well on that front.


Sure, and VW also self-reported that their engines passed emissions testing and Ford self-reported that the Pinto was safe. Self-reported numbers by the manufacturer who has a direct financial interest to lie and who maliciously miscategorizes their L4 design intent system as a L2 design intent so that they do not need to do safety testing or submit reports to regulatory agencys is the opposite of believable.

Except this leak confirms them in this case hence the subject of this thread?

  > My model 3 does something scary at least a few times a month but I've never reported any.
Why not? The barrier of entry is pressing a button and then describing, in voice, what just happened.

The privacy violation of doing so and communicating who the hell knows what metadata to a man-child who will happily disclose it if he feels like it one day?

That's a fair point.

Not least because this is why I wouldn't consider buying a Tesla. It feels too much like dealing with software: it isn't your car. It is Teslas car. And maybe paying a bunch of money for it means you can drive it, but sometimes, when you hit some corner case, it doesn't.

As a Tesla owner, I would love to know what corner cases you are referring to.

I've been a tech at a Ford dealership, and I'd say that the incidents which render the average Ford undrivable are far greater than the incidents which render the average Tesla undrivable.

Second hand Teslas represent a challenge as someone pointed out. You also can't necessarily modify or inspect the car without Tesla getting their panties in a twist (I dare you to connect to the network inside your car and try to get access to the onboard computers and then try to have the discussion with Tesla about who's car it really is).

The there are the massive privacy issues and the fact that you can't do anything about that: your car is spying on you. (I know consumers don't care enough about this in the US to elect politicians that take privacy seriously, but in Europe we do. Which is why Tesla is under scrutiny for not only collecting massive amounts of invasive data, but also for not being able to adequately safeguard the data they collect).

I'm not in the US. My Tesla has an option to disable data collection. You could argue that the company might not honour that option, but that argument then opens a very broad subject across many industries.

This is a much hairier and more nuanced conversation for a different thread, but second-owner (used) FSD comes to mind as an example of not truly owning the car.

> Why not?


No, seriously. Why?

Most people are (correctly) disillusioned enough to understand not to use a company-provided feedback funnel.

It is possible that this one is different. Doesn't change the fact the correct first impression is to never use the feature.

> Why? > No, seriously. Why?

To improve the feature. The same reason that we report bugs to KDE, Mozilla, LibreOffice, Canonical, Debian, Red Hat, Anki, JetBrains, Oracle, etc.

I submit bugs to open source projects as they are transparent and it helps developers.

I don’t submit random bugs to Oracle because nothing happens unless it’s serious and I have paid for support to have dedicated engineers debugging stuff.

yea, in oss projects, users tend to feel more like stakeholders than with products of some mega-corp.

we all know that these feedback channels just create a corpus of opinions to be used for arbitrary arguments by management. no one is gonna read or even react to this unless it's necessary.

I get this a lot with institutional support at my organization’s IT. “Why didn’t you report it, it just takes a minute.”

The effort is in the follow up time and the pointlessness.

Submitting an IT ticket is different than a Tesla event. But I don’t want someone to ask me if I’ve rebooted my computer and trying to call and confirm I’ve tried all the things I said I tried.

I never submit tickets unless I expect some strong probability of solution.

There's no reason to assume that of owners of other car brands are reporting 100% of issues, so relative comparison is still valid.

Why not?

I was going to say "Because you believe in the Tesla mission. That's why someone buys the car." but I realize that that's actually no longer the case. That they've become a car manufacturer where customers can simply weigh the pros/cons as they decide what car to buy.

my six year old model 3 with 60,000 miles on it doesn’t do anything I don’t tell it to- this sounds like user error. I wouldn’t report user error either.

It’s not the burden of society to prove the dangers of the technology we purchase on behalf of the businesses that sell it.

We recall consumer products for danger well below the numbers reported in the article above. A vehicle on the road should be held to an even higher standard.

> We recall consumer products for danger well below the numbers reported

Aren't recalls related to systematic failure and severity, instead of absolute number of occurrences of an issue?

It can be both.

Safety testing has proven that Teslas are incredibly safe - probably some of the safest cars on the road. Both in crash tests and accidents per mile driven.

If someone is claiming that there is a safety issue despite the existing data, then yes, the burden of proof moves to them to validate their new claim.

Their claim IS the data.

Except that the data leaked does not demonstrate that the Teslas are unsafe.

No, of course not, a company with a policy to sweep accident data under the rug can't possible be unsafe, can it?

Do we? I don’t think so. We recall vehicles based on an investigation and quantification of the risk.

The numbers here aren’t very high and they don’t reflect any serious risk of harm. Recall math is interesting to me, but it’s not just a function of customer complaints.

I am wary of the word “society” when it means in this case “a counsel of my design”, but otherwise agree with the sentiment.

Do you know how many sudden unintended acceleration were needed for Toyota to take a beating? How many fires were needed for Ford to recall Pinto? How many ignition failed for GM to take a beating?

All of those were low numbers.

No I do not. What were those nummbers?

Toyota recalled 9 million cars for about 40 deaths.

Ford recalled 1.5 million cars for about 40 fires.

GM recalled 30 million cars for about 120 deaths.

To put things into perspective, about 100 people die from traffic accidents each day in USA.

In Germany alone and for a period of time.

No. Despite the data being leaked to German newspaper most cases are reportedly from the US.

The overall safety numbers for Teslas speak for themselves. Teslas are some of the safest cars to own.

Obviously these issues should be solved, but net-net, Teslas are very very safe.

How different is this from any other similar sized companies behaviour?

This site doesn't seem that impartial/unbiased. From a linked article on Tesla:

> It’s also worth noting that the above email was sent after 2 a.m. Pacific, which isn’t specifically relevant to the faked video. But it does make it look like Musk is a loser with no friends or anything else to do other than work. Loser.

When a Toyota hid an “unintended acceleration” bug, it was a scandal that resulted in a $1.2B fine.


Every customer complaint about unintended acceleration of Teslas was proven in court to be the fault of the customer confusing the pedals. The accelerator pedal has two independent sensors measuring input - they both have to agree, and no input on the brake pedal must be detected for the vehicle to accelerate.

They go a step further. They use their cameras to detect the environment and significantly slow down acceleration if they think it might be a mistake.

It was the same with Toyota too. Car and Driver did an article and summarized the unintended acceleration cases and it turned out most people were intoxicated. Same thing with Audi back in the early 90s. People were pressing the wrong pedals and blaming the car. Audi still lost and Toyota still lost.

Two things can be true at the same time.

1. Most unintended acceleration events are user error across all vehicles. 2. Some small number of Toyota unintended acceleration events were caused by low quality software on the ECU and a very badly implemented watchdog that did not implement brake override or other safety measures.

No such thing was proven in court.

Instead, what was shown in court was that Toyota had a culture of firmware engineering that produced code impossible to consistently QC, debug, test or verify. And as a result, they quietly fired the directors of that department, rebuilt it from scratch, and replaced every TCU from that era with a re-engineered unit in a series of about a dozen recalls spanning a decade and millions of vehicles.

I think you confused toyota with tesla

you're right, thanks for pointing that out... got confused by another reply in this thread, and whenever I see "unintended acceleration" I think "toyota" and the names kinda look alike. Too late to edit my comment.

Teslas have a problem with phantom braking (e.g. when there's a dark shadow it fails to detect that it's a shadow before going from ~70mph to ~55mph). Myself and countless friends I know have experienced it but that problem has yet to be solved.

A lot of cars with automated emergency braking have problems with phantom braking. This is an elephant in a room that hasn't been opened yet, but eventually will be.

Like others have said, this would be less of a problem if Tesla relied on LIDAR like others. Tesla uniquely has this problem that's unsolveable because it's relying on cameras only.

This lidar cult is getting ridiculous lmao how did you people come to identify with a sensor so much

Next to no cars use lidar for driver assistance. The vast majority uses MobilEye.

I drive a Tesla day to day (with the latest FSD Beta) - yes, phantom breaking was real - but it's been almost completely remedied for the past few months in my opinion - and I'd be shocked if other automakers (as was mentioned above also have these problems) are as quick to iterate based on feedback.

others don't use cameras :D

Autonomous driving companies are using LiDAR but almost every consumer passenger vehicle is relying on a vision based system, and they all have this problem. Look behind the windshield of every car you walk past :D

Would you have a link to that? I'd be interested to look.

As a side note, I've never heard of confusing the pedals as an issue for ANY car, so if Telsa's get people to confuse them enough to bring them to court, it's probably bad design.

Edit: The US National Highway Traffic Safety Administration estimates 16,000 accidents per year in the United States occur when drivers intend to apply the brake but mistakenly apply the accelerator.[3] from wikipedia on Sudden unintended acceleration

Hard to imagine how you fail to design pedals!

> As a side note, I've never heard of confusing the pedals as an issue for ANY car

Then you haven't been paying attention around the Toyota acceleration scandal.

Pedal confusion is remarkably common when you buy a new car/use a rental (your feet rely on muscle memory), and it's not uncommon in the elderly.

Most of the Toyota unintended acceleration fits the statistical profile of pedal confusion in the elderly.

However, what Toyota really got whacked for is that when people pulled their software for audit, the software was a disaster and didn't even adhere to basic standards. At that point, it was cheaper for Toyota to just admit fault than go through with a whole lot of court cases that they were likely to lose once a jury got involved.

Oh wow! Thanks for the heads up.

When skimming about Toyota, I'm getting unsafe floor mats and sticky pedals as the cause of acceleration, but maybe I'm not looking hard enough. The other commenter also brought up that it's a common issue.

Guess I'm feeling less safe on the road then ever - and I'll get a manual to boot

If memory serves the main ECU control loop didn't check for stack overflows so excessive recursion could smash the global variables on the stack and accidentally turn off any number of ECU tasks - including the one responsible for monitoring the accelerator and brake.

The ECU module includes watchdog support that runs on another chip or core (can't recall) that was intended to do backup monitoring of the main ECU - and especially it should have watched the brake pedal and of the brake was held for a minimum time it would override the ECU and force the accelerator to zero. However that function did not work reliably, making the watchdog useless.

The code itself was poorly structured, with lots of critical things done in one big "god" task that if accidentally disabled by a single bool flip in RAM would ultimately disable many safety critical functions. Normally you'd have multiple copies of such data structures that must agree, split the code up into separate isolated tasks so a failure of one doesn't stop the others, and implement basic stack overflow protection which again IIRC was available on the toolchain they were using but was not enabled.

The watchdog problems are especially inexcusable for a safety critical system.

Another one crash that I think about semi often was a bus crash in Queens that killed 3 people where the bus was doing 60 in a 30 and blew a red light. The cause was suspected to be a thermos wedged under the brake pedal and pushing on the gas. It seems like a dumb thing to think about sometimes, but I try to be aware of that risk. https://www.core77.com/posts/84534/How-a-Commuter-Thermos-Ma... https://www.nydailynews.com/new-york/queens/ny-metro-dropped...

I've definitely had a few instances (over a few decades of driving) when I lost confidence in my knowledge of which pedal was which. Rote knowledge is tricky that way. Fortunately, I was always able to safely test. I was never confidently incorrect, but I can see it from here. It's a scary thought.

Friend of mine works for a different well known car maker and tells me the software is exactly like that.

How many people will have to die before governments impose the software to be submitted for approval?

It was famously a problem for Audi in the 1980’s and almost destroyed the brand. I thought everyone had heard of that.

But it is one of the most common if not the most common cause of unintended acceleration in any car.

I’ve even had it happen to me one time. The typical scenario is that you’re traveling at low (“creep”) speed with your foot on (but not pressing) the gas pedal. You think your foot is on the brake, so you push to slow down… whoops you’re starting to accelerate.

The probable reason that it happens more often with Teslas is that they have less lag between pressing the accelerator and getting juice. So by the time you realize you messed up, you’re already going fast.

In most gas cars, firmly pressing the accelerator results in milquetoast acceleration and a lot of noise for a second while the transmission downshifts and the engine revs up. In an EV, you just… go.

This is why I brake with my left foot and go with my right. Cuts the latency and confusion to a fraction.

> As a side note, I've never heard of confusing the pedals as an issue for ANY car, so if Telsa's get people to confuse them enough to bring them to court, it's probably bad design.

I think this is media bias. The media picks up accidents involving Teslas far more often than they do other manufacturers. The national news will even cover Tesla recalls when it's just an over-the-air software patch with zero known real-world impact), and similarly despite that there are ~25K vehicle fires per year, you only see them in the media when a Tesla is involved. In particular, confusing pedals is pretty common, particularly among very old drivers.

>confusing the pedals as an issue for ANY car,

I remember NASA thought so. So I just posted it, thanks for the post idea!


It happens all the time. People panic and mash their foot to the floor. Some old guy crashed into 6 cars down the street from my work through a parking lot, and jumped his car across the alley and though the wall of my office into our break room. He just panicked and thought he was mashing the brakes. Luckily nobody was in the break room.

I think it's more that if someone does it in a Tesla, it makes for good headlines and generates clicks, so we hear about it. Someone in my hometown confused the pedals in an ICE car a few years ago, made a small blurb in our tiny newspaper, nowhere else. Same with cars catching on fire. Happens all the time, but when it happens to a Tesla, you see it in the national news.

I googled driver confuses pedals and got a ton of results. Surprisingly it happens pretty often. Makes me a little nervous.

> Every customer complaint about unintended acceleration of Teslas was proven in court to be the fault of the customer confusing the pedals.

Reference, please? Are you referring to a specific complaint, or truly "every" customer complaint?

I was kinda scared, or at least on edge, during the Toyota stuff; I usually have a very level head about this stuff but for some reason that one got me. But if it happened again, I wouldn't really blink an eye.

Is what you raise relevant? Is it fine if everybody does it? Is the leak not real because you found a link to it on an unimpartial site?


The leak can be real, but the narrative around the leak is important.

My read from this link to a story about a story:

~1000 crashes related to autopilot reported for the 2.6 million autopilot enabled vehicles shipped in the reported time frame.

.04% total failures. How many of these were user created and not the fault of the car? How many of the failures from the car were specific to that cars hardware vs the software? Was it a Tesla hardware failure or an OEM device failure?

I’m not gonna do a full analysis, but whenever I re contextualize myself on car crash statistics I am reminded that Tesla failures represent an insignificant fraction of all failures.

I think you meant 'biased' and not 'impartial'. I am not being pedantic; I don't want your sentence to have the wrong meaning so I am bringing it to your attention if that is the case.

Oh no thanks! it was a typo, added un at the begining.

The report being on a site with a bias does make me want to read with a more critical eye, as everything is being spun to look as bad as possible.

The behaviour described (not wanting anything written) might very well be standard operating procedure with any company with decent legal counsel. If it is common behaviour, it is not some sort of Muskism, part of his evil scheming as is implied in the article, instead it is a reflection of the world we live in.

If you can find the same information to corroborate then no, but if it's the only source...

But it is one of those trends that I personally do not like either where these posts are made like it's a personal conversation between two girlfriends or whatever. There's opinion pieces and then there's this kind of I don't even know what it's called I'm so un-hip

What's stopping you from clicking through to the German source and reading the English version of their piece?

Here's the German article in case anyone wants to read through it / translate


Jalopnik is absolutely garbage "journalism", they have a long long history of hating on Tesla.

They only refer to a respected German business newspaper which seems to have reasonable journalism.

That doesn't mean that those journalists are necessarily completely unbiased. Germany's traditional car industry is threatened by Tesla without any doubt.

However, the details on how Tesla is handling these reports is revealing and I have no doubts that they are real. That those who publish them might not like Tesla does not change that.

Isn’t it a gawker site? I wouldn’t expect anything less.

> How did the company deal with complaints? The Tesla files also provide information about this. The files show that employees have precise guidelines for communicating with customers. The top priority is obviously: offer as little attack surface as possible.

For each incident there are bullet points for the “technical review”. The employees who enter this review into the system regularly make it clear that the report is “for internal use only”. Each entry also contains a note in bold type that information, if at all, may only be passed on “VERBALLY to the customer”.

“Do not copy and paste the report below into an email, text message, or leave it in a voicemail to the customer,” it said. Vehicle data should also not be released without permission. If, despite the advice, “an involvement of a lawyer cannot be prevented”, this must be recorded.

Customers that Handelsblatt spoke to have the impression that Tesla employees avoid written communication. “They never sent emails, everything was always verbal,” says the doctor from California, whose Tesla said it accelerated on its own in the fall of 2021 and crashed into two concrete pillars.

"Is your data also in the Tesla files?" (German ) - https://www.handelsblatt.com/unternehmen/industrie/leseraufr...

Its my understanding that the vendor that they were buying LIDAR from was either slow paid or never paid - and then cut them off. This is when LIDAR became less important in their marketing spin and they said that they do not need it.

I didn't see the actual data or a link or anything to how to get it, but why on earth would it be 100GB? That seems fairly large if it's just tallying accidents and such.

"The 'Tesla Files' comprise more than 23,000 files. Some documents apparently show salaries and home addresses of more than 100,000 current and former employees. Others list presumably private mail addresses and phone numbers of customers."


"Due to the sheer volume and structure of the data sets, the editorial team had to make a selection of which of the more than thousand Excel tables could be taken into account in the query tool. "


If just a few of the accident reports have attached photos or videos, you'll quickly get to 100GB.

Tesla seems to be making progress on full self driving, but real self driving (cars without steering wheels) still seems quite distant. How will such a car respond to road workers redirecting traffic? How many times have you had to talk to someone outside of your car to obtain instructions on how to get around some obstacle, like a moving truck. I don't do such things every day or even every month, but there are occasions where it is necessary to take unmarked detours.

How would a self driving car get through heavy fog or snow covered roads when there are only very difficult to decipher hints to the road edges?

How will self-driving cars deal with humans that can perfectly predict their actions? I believe that bad drivers will take advantage of self-driving cars by cutting them off and failing to yield when they should.

Because human drivers have a sense of self preservation we can break the rules when we are about to be car-jacked or see an impending collision coming. Imagine how easy it is to obstruct a self-driving car for nefarious reasons.

I'm confident that, eventually, self-driving cars will address all of these issues to some degree, and that will mark a turning point where it is better to leave driving to the cars than average drivers. However, this doesn't mean I'm going to put my $100K car into a taxi pool to be used by (sometimes drunk) strangers; I don't buy the argument that the cost of buying such a car will pay for itself.

German news site is a bit misleading. Handelsblatt is the leading business newspaper in Germany. And according to the article they made significant journalist effort to confirm the authenticity of the information.

Unfortunate thing for Tesla is we have the Affective Fallacy where if we like Tesla we overweight its benefits and underweight its errors.

What's the name for that fallacy where we do not know what went right, and we pessimize to assume the worst? I.E. How many times did Tesla prevent a crash? Assume none?

Don't know that it's a fallacy, it's more that the question you ask is impossible to answer - it's impossible to have a number for how many crashes were prevented. But we do have a number for crashes caused by. So we work off of that - if Telsa want's to argue they are safer, I'd love for them to devise some way to prove it, but I can't see that happening.

Forbes did this with stats. They concluded autopilot was not causing accidents.


It seems like comparing AP to anything but regular cruise control miles is not very fair and is going to inflate its safety.

From that link, it's pretty bad if it's not even clearly better at general "highway miles" than when it's off because people are least likely to use it at the times that are much more dangerous generally, such as merging onto and off the freeway.

Surely you're familiar with this? https://youtu.be/3mnG_Gbxf_w?t=7

You can figure out statistically if Teslas are systematically avoiding some types of accidents.

I think you can still try to be objective.

For example, I like tesla autopilot 1 (mobileye based) and found its behavior to be reasonable. I also know its limitations - it has only one camera - so I don't use it in edge-case scenarios.

I think model S display is good for turning settings on and off, but is bad for controlling any aspect of the car especially critical ones like headlights, climate control, door locks and more.

I think the newer model S/X cars are bad, replacing stalks with touch controls for the headlights, turn signals, horn and wipers. They get in the way of good driving.

After driving a newer model I have to disagree. I find it more efficient and natural, not sure why people stress over it. Getting used to yoke was way quicker than regenerative breaking.

Why precisely makes that a fallacy?

You're presupposing the conclusion.

If you didn't overemphasize the good parts and downplay the bad parts you might find out that maybe you shouldn't really like what Tesla is doing at all, or you'd have a legitimate reason to like Tesla.

What you shouldn't do is overemphasize the good parts so you can feel better about your decision to like Tesla. Then your conclusion has been fixed from the start and you're adjusting the facts to fit to avoid cognitive dissonance.

When does a thief become a whistleblower?

It seems to me that something really dangerous or illegal needs to be going on to morally justify publicly releasing a 100GB worth of (presumably unfiltered) private company data.

The data also contains salaries information and private addresses of Tesla employees, which clearly don’t belong to the category of “whistler blower data”, even under EU laws.

Source article text, Handelsblatt: https://pastebin.com/S49ttUqf

Jalopnik has been very negative in Tesla articles for several months now. (I read auto news all the time, I love cars.)

I wonder if they're acting on behalf of another auto manufacturer.

And this the real risk of alienating a group, they’ll retaliate when they no longer feel aligned with you. Even if they’ll work for you, they won’t have the same allegiance/loyalty they had before.

3k complaints for 3 million cars? "Only verbal" (common practice)?

Sounds like click bait more than anything.

Garbage technology by a company with bunions backers.

Can someone link to the source article? It is paywalled

Horrible company, with a miserable CEO.

I avoid driving anywhere near Tesla’s, especially tailing them. Phantom braking is not okay.

You shouldn't tail any car in general. There's a lot more 'real' braking out there than 'phantom' braking.

in most commutes, if your tailing distance is too much, someone will else will come in front of you.

if you think about the speeds involved, a single additional car in front of you on the freeway (or even any additional cars) adds pretty miniscule time to the total commute.

Let's compare a few situations. In the baseline you're tailing the car in front of you with a focus on not letting anyone cheat and get in front of you, let's say 50 feet away. Your commute is 30 miles, and in this frictionless sphere of traffic you're going 60mph the whole time. You get to work in 30 minutes flat.

In the second scenario you're following the 3-second rule[0]. This would put you ~285 feet behind the car in front of you. Let's say over the course of your commute 20 cars move in front of you. If the average car length is 15 feet, and they all are 50 feet away from each other, when all 20 cars are in place you're a net -(20 * 65) feet away from the original car, or 1300 feet total. At 60 mph that adds ~15 seconds to your total commute time.

Well worth having an easier time avoiding a potential crash IMO! Also has the benefit of helping prevent traffic to begin with[1]

0: https://driversed.com/trending/what-safe-following-distance.

1: https://www.youtube.com/watch?v=iHzzSao6ypE

It is not about efficiency or trying not to be slow. I am not bothered how many people cut in line or "cheat".[1]

Someone will always keep cutting into the space. It is impossible to maintain 3 second to the car ahead. First one car will cut in, you make room for them and add 3 second gap the next car will cut in. Maintaining even a 1 second gap is sometimes very hard close to exits.

[1] Personally my driving pattern changed once I switched to driving a hydrogen fuel cell Mirai, slower is better on the mileage and fuel cell owners are very range conscious.

it's anecdata but I drive this this on the regular and don't have any issues! May be some regional differences

In the Midwest here and drive the same way, from rural to busy city and back. Never had any issues at all with it minus the car or five I let in front of me. I usually keep about 5 car lengths, so a fairly reasonable gap without being annoying to drivers that happen to be behind me. Also, I have zero panic braking incidents, which seem to be a big cause of crashes and slowdowns in heavy traffic due to the slinky phenomenon that tends to happen.

Not only can I see everything beyond the car in front of me, I tend to "soak" up the braking energy when there is a panic braking incident in front of me. I have no data to back this up, but I'm almost positive I've kept traffic moving much better behind me and prevented rear ending incidents using this tactic. Also, I wouldn't downplay the amount of fuel that is being saved that comes with not having to almost stop, then start over again from the slinky effect.

Tell me you're a bad driver without telling me you're a bad driver.

You need to not care about that, and you're actually supposed to let people change lanes into your lane. You're getting into a mental competition with other drivers and sacrificing the safety of yourself and everyone around you.

And if anyone believes that longer following distance causes more traffic, that is also false and the reverse is actually true. It is the poor reaction times of tailgaters that cause traffic slowdowns.

> Tell me you're a bad driver without telling me you're a bad driver.

> You need to not care about that, and you're actually supposed to let people change lanes into your lane. You're getting into a mental competition with other drivers and sacrificing the safety of yourself and everyone around you.

> And if anyone believes that longer following distance causes more traffic, that is also false and the reverse is actually true. It is the poor reaction times of tailgaters that cause traffic slowdowns.

I think that what they're saying is the flow of cars in front of them keeps that distance between them and the 'next car' to a shorter undesirable distance as more cars fill that gap during traffic.

... And then you end up just going slower than the rest of the traffic, and people behind you change to the faster lanes to pass you, and some of the people passing you change lanes back to in front of you. And so trying to keep a longer following distance than the rest of traffic allows just means thtey lots of people are doing things other than just staying safely in one lane.

you're better off letting people change lanes to the lane they want to be in, and if you're not camping in the left lane then typically you'll find people changing lanes from the left lane to exit, or will change lanes through yours into the left lane.

if you leave lots of space they can do so entirely safely.

trying to prevent other people from changing lanes does not enhance traffic safety.

then you let off the gas slightly and widen the gap. when people change lanes from in front of you, then you can speed up slightly and close the gap. that isn't a practical problem. it comes out in the wash.

That's fine. You lose like .5 seconds of your life when another car comes in front of you.

It doesn't stop with one, the problem is not someone cut across, the problem is someone always keeps squeezing in the gap. Every time you open a gap, a new car cuts across. How can you maintain any gap in that context ?

As others pointed out - let them. On average all lanes move the same, they might even move away once the lane stops moving. You'll have much worse time rearing a car then letting all those cars in.

I also keep additional distance in traffic to minimize slowdowns/stops, which ultimately actually improves/fixes the flow.

So? Drop back again.

Then another car will do that… and then another… and then another… in the end it becomes way more dangerous.

So what.

That's not what OP was implying.

Real braking happens for a reason, to avert accidents. Sane drivers don't do it just randomly in perfectly safe situations. Random braking is major risk for everyone around and if you are source of it, you are in fact danger foe others.

> Sane drivers don't do it just randomly in perfectly safe situations

You can't always assume that you will be able to tell when a situation has turned from safe to unsafe. You just can't exactly see what the driver in front of you (or the driver in front of _them_) is seeing.

And you can't assume that the driver in front of you is sane!

You have to _always_ follow at a safe distance.

Humans in traffic are relatively easy to predict.

They don’t just randomly brake for no reason.

But sometimes there are legitimate reasons for a car to brake hard, which you'd never know from your position one car-length behind them.

I hear people talk about people brake checking all the time (whether they're the brake checkers or the person tailing). A buddy of mine was tailing a (non-Tesla) car a little too closely on his motorcycle and the driver deliberately brake-checked him and he wrecked. And that's just people deliberately driving erratically, never mind the people who are responding to debris, animals, people, etc darting into the road.

Fault of your friend for not keeping a safe distance.

If both drivers only had basic insurance covering damages that aren't your fault, he got nothing and the other driver got paid for having his car damaged by a reckless driver that smashed into his car.

How does your friend even know it was without reason? Maybe a kitten or a squirrel crossed the road.

We’re not debating fault, we’re debating whether other drivers drive predictably. Whether the driver brake checked him or whether there was a squirrel or kitten in the road doesn’t really matter: the point is the human driver braked unpredictably. You don’t have to drive with special caution around a Tesla, you should drive with the same amount of caution that you would drive around any car. I have a Tesla and I use autopilot pretty liberally (but not carelessly—I’m always ready to take over) and phantom braking was really frustrating last year, but it was never unsafe unless perhaps someone was literally a foot or two behind you at high speed. They seem to have fixed that sometime in 2022, and since then I’ve had fewer than 5 phantom braking episodes and they have been almost imperceptible (certainly not dangerous). The Tesla critics in this thread are reacting to some pretty extreme assumptions about autopilot (that phantom braking happens all the time, that it’s less safe than the unpredictable behavior of human drivers, etc). We certainly need more evidence, but I’m going to be biased toward the people who actually have experience driving Teslas rather than the people whose information comes from the rumor mill.


Sure, but an unexpected full ABS lock when you can see nothing in front of the tesla is going to be yard for most to react quickly too even at a decent distance. 10 second follow distances are only possible in most metros during very light traffic.

Then you're driving unsafely, although 10 seconds as a general rule is a straw man.

At any time, a child could run out from a hidden spot and the car in front of you could have to slam on the brakes as hard as possible. Or any of a hundred other realistic scenarios.

These things aren't common, but statistically they will happen to you multiple times during a lifetime of driving, and it's your responsibility to always be at a safe distance behind in order to react as well.

The common rule of thumb is generally 2-3 seconds in perfect conditions, and 4-6 seconds in rain or other normal challenging conditions. 10 seconds is only in cases of ice/snow where most people wouldn't be driving in the first place (you know, when you're going just 15 mph but it still takes 5 seconds to come to a full stop on the slippery ice). The heaviness/lightness of traffic is irrelevant.

Phantom braking is not a full ABS lock type of braking. It’s more like a brake check.

>10 second follow distances are only possible in most metros during very light traffic.

I've usually heard that it's three seconds. Even still, you control your follow distance. Even in heavy traffic, you can give yourself more space between you and the car in front of you than other people do. It's easy to do, and I've been able to do just that even in metro areas with heavy traffic.

You only need 2 seconds to be safe. It's longer than you think.

3 is the guidance in the UK at least

> You should ... allow at least a two-second gap between you and the vehicle in front on high-speed roads and in tunnels where visibility is reduced. The gap should be at least doubled on wet roads and up to ten times greater on icy roads

Highway code rule 126[0] says 2?

[0] https://www.gov.uk/guidance/the-highway-code/general-rules-t...

Oh. I stand corrected.

I mean, I think 3 is better, especially at high speeds. 2 is just the minimum.

I avoid them too. Erratic drivers too. I'm not sure if its them coming out of autopilot back into control or just that's what happens when you have a lead foot and a car that goes to 60mph in 3 seconds entirely silently.

I treat them like semi-trucks on the highway. Pass them or let them pass you, but don't loiter alongside, behind or in front of one.

Why are you loitering beside other cars? This is a good safety practice regardless of the car’s make and model.

It's a shame this is getting downvoted because it's true. In any level of traffic except stop-and-go--and even then to a large extent--you should be doing everything within your power to either pass the cars on your right, and/or move over to let cars pass you on the left. It is a vanishingly small number of scenarios where you are keeping pace with a car next to you and you're not in the wrong. This isn't just at high speed or on highways, any two-land road operates (or is designed to operate) this way.

That's not the same everywhere in the world though, and even in places where it is strictly passing on the left (or on the right in the UK, Japan and a few other places) 'keep your lane' tends to be the rule if the right hand lane is also moving at the speed limit (so you can't legally pass).

That way the carrying capacity of the road is higher. But when traffic is less dense 'station keeping' should be avoided at all times and if someone moves into my 'dead zone' or just to the left of me I'll gradually slow down to force them to finish their overtake.

> 'keep your lane' tends to be the rule if the right hand lane is also moving at the speed limit (so you can't legally pass).

Why are you in the lefter lane if you can't pass?

I've been in rush hour (where keep-to-the-right-unless-passing is very strictly enforced) in bumper-to-bumper traffic and the left two (out of 6) are completely empty and everyone is doing 'around' the speed limit. Some are in the right lane doing a few below the limit, some are in the left-most lane doing a few above.

Occasionally, someone who is late to work, emergency services, or whatever goes flying by in one of the left-most lanes.

> I've been in rush hour (where keep-to-the-right-unless-passing is very strictly enforced) in bumper-to-bumper traffic and the left two (out of 6) are completely empty and everyone is doing 'around' the speed limit.

I've been on interstates in every continental US state and I've never seen this, but I think something has been lost in translation because "bumper-to-bumper" and "everybody doing the speed limit" are mutually exclusive as I understand the terms. If everybody on the road can fit into the right lane with enough space in-between to do the speed limit, that is done but I wouldn't call that traffic "bumper-to-bumper". I would call that light traffic. Bumper-to-bumper is when the space between cars really starts to contract, because everybody is going substantially below the limit, or because people aren't maintaining a safe distance.

Once the road has too many cars to fit them all into the right lane at the speed limit, then in every state I've driven, cars start using the left lane for travel, not just passing. If the right lane is so full that it can only sustain 5 below the limit, then people start driving in the left lane and stay there for as long as the right lane won't support speed-limit traffic. In this kind of traffic you'll start to have cars moving fast alongside each other with low relative velocity.

Have you ever driven in Philadelphia during high-traffic, non-rush hour times? 75mph easily, with at most half a car length between every car. I didn't believe a friend's dad when he talked about "bumper-to-bumper 80mph traffic" in the highways around the Philly suburbs but it's absolutely the case.

We're talking 1-2 meters between your car and the car behind/front of you. It's extremely stressful to drive in at 70 km/h.

!!!! 2 meters separation at 70 km/h gives you a tenth of a second to react to anything the car in front of you does, that's flatly insane. Where in the world do people drive like that?

Seriously, that's objectively insane. Try the ruler drop test for reaction times if you don't believe me, a 10th of a second to even initiate your response isn't realistic and obviously gives no time for the response itself to have effect. What I'm saying is that at a tenth of a second, you can't even start to press the brake pedal in time, let alone have enough time to actually slow down.

In America, with only 2 meters between vehicles the traffic would be inching forward at a snails pace, under 20 km/h at least. "Stop and go", as in people would stop their car and then drive forward slowly when a larger gap ahead of them appears.

If the car in front of you instantaneously goes to 0, yes.

Reminds me of driving in western Virginia, people drive insane there. This is going into Amsterdam in the morning. Highly recommend “Not Just Bikes” on YouTube if you want an interesting comparison between roads here vs. Canada/US.

Where is this? In eastern canada that's impossible to imagine - although most highways are 2-3 lanes, not 6, I couldn't imagine having a free lane on the side while having bumper to bumper everywhere else.

Or are these protected lanes for carpooling?

I find this very impressive!

> It is a vanishingly small number of scenarios where you are keeping pace with a car next to you and you're not in the wrong.

I think it really depends on where/when you're driving. I find this to be a common scenario on interstates during rush hour:

I'm in the right lane, doing approximately the speed limit. There is a safe distance between me and the cars in front and back of me, but only just. If many more cars enter the road, traffic would need to slow down to maintain safe distances. In the left lane is the same situation, except they're averaging about 1 or 2 mph faster. In this situation, there are cars in the left lane passing very slowly, spending a lot of time alongside me. I could slow down below the speed limit every time a car passed on the left, to reduce reduce that loiter time. But this would make my driving less predictable to the drivers behind me (and waste a lot of mileage too...)

So normally, when the other cars are my size, I maintain my present course and speed, driving as predictably as possible to help the other drivers anticipate my course. Changing position in traffic is inherently risky, so I avoid making changes unless doing so is necessary to avoid something I judge to be more dangerous than the average. If a truck passes me on the left, I'll slow down to make the passing faster even if that means a car behind me has to brake. But if in that moment I judge the guy behind me to be even more dangerous, then maybe I won't. It's the kind of decision that needs to be made on the spot in a case-by-case basis. On interstates that are flowing fast near capacity, you need to be constantly evaluating the relative threat of the traffic around you.

I think if you're in the left lane only going 1 mph faster than the cars in the right lane, especially in traffic, something is wrong. Especially on interstates where lanes are typically wide and the road relatively straight, and in good weather & visibility, if you have any appreciable amount of traffic on the road it's much better to go 5-10 over to pass then get over to the right and go back down to the speed of the right lane than it is to be in the left lane holding up a line of traffic while you take 5 minutes to pass one semi.

In light traffic, I wouldn't. In medium traffic there often isn't that much choice, but in those situations I prefer the company of other cars my size, and preferably ones with attentive drivers (so I discriminate against Tesla drivers.)

Most people are pretty much unaware of anything outside of their car other than for the couple of seconds they look up from their phone to look at the car in front of them. Look to the sides? That's too much time away from the screen in their hand! /s (only partially)

Since my time of learning to drive, the requirement to have formal driving training has ping ponged in being a requirement or not. The number of hours as an observer is just as important as the hours being behind the wheel. One of the things repeatedly mentioned by the instructor was to not drive side by side any car unless absolutely necessary. It was also a recurring theme in my repeated defensive driving classes. I also have an uncle that drove trucks for a long time, and he would tell stories of things he saw on the road. A relevant story was when one of the wheels of a tractor-trailor doing 70mph down the highway lost the outside wheel of the trailer and seeing the damage it cause the car driving along side. All of that added together makes me never like to have a car on my sides and I will speed up or slow down (which ever has more space available) to avoid it. For those that did not have to take a driving course, this is just information they may never have been provided.

It is totally normal to just go behind one car without constantly overtaking or being overtaken. It is even actually safer then being constantly in and out of lanes.

I treat them like the safest cars to be around. Because statistically, they are?

According to statistics that Tesla publish.

I would like to see mercedes publish their own statistics on their FSD. Given the tight constraints where they allow self driving, they could easily claim 100% safety record and thus infinitive more safe than any other manufacturer. It would be misleading, but statistically it would be the truth (any accidents could be said to be outside the constraint and thus will not count).

Personally I would only really trust such statistics if insurance companies would reflect that in the premium. Somehow I doubt they would be willing to cut the fee based on what Tesla claims.

Can someone present evidence that Teslas are more expensive to insured than other comparable vehicles? Along with evidence that the difference is because of safety rather than cost of repairs?

Tesla has their own insurance because people were complaining about high rates. Why else would they come up with their own insurance?

> Why else would they come up with their own insurance?

If they had evidence that their cars were less likely to be involved in accidents then it would pay them to set up an insurance company and charge just below the usual rates.

Even if they are merely as safe as other cars then Tesla could provide insurance at lower prices because they need only break even and they have better statistics.

That's not proven as far as I know, unless you've got stats? (And not the stats from their website - those are not comparable)

> And not the stats from their website

There is a reason it is said, "There are lies, damned lies, and statistics." Statistically, there's a ~7% chance I won't die because ~7% of all humans who have ever lived haven't died.

Citation needed, Musk’s tweet doesn’t qualify.

Hope this link posted further down the thread does qualify though.


That's not a great idea. The driver is still in control of the car and is no better than anybody else, and maybe even not paying attentnion because some think autopilot is self driving. You should treat all other traffic with the same defense.

if you are on the inside at least... But yeah i think they tend to have better track records than human drivers.

There's a difference between safety and consistency.

Agreed. With human drivers you can, in most cases, see their intent with the way they move. With tesla or other self driving systems it is a black box that is way less predictable.

sadly this is virtually impossible in Los Angeles.

The Tesla drivers in LA are like an upper echelon of terrible Tesla drivers:


phantom breaking isnt a thing in 2023. it used to be a problem, it's not anymore if you have HW3

Spoken by someone who clearly doesn’t own or drive a Tesla. Phantom braking is very much STILL an unsolved problem.

I own a tesla, im on fsd 14.2 though. maybe it's still an issue on auto-pilot. i have no had an issue with phantom breaking in over a year.

This is not true, I purchased a Model 3 brand new in 2021; it has HW3 and it absolutely does phantom braking regularly on road trips. This is on US interstates.

Why do you keep driving it? I had a car that did this to me twice and got rid of it, the first time could have been a glitch but twice is simply broken and dangerous. The dealer said the car was fine.

All car brands ADAS are imperfect. If it phantom brakes you just push the accelerator, or simply don't use the ADAS.

You can drive the car normally, you aren't forced to use ADAS if you don't like it.

Yeah just gotta turn it off every single time you get in the car.

No? What are you talking about. Just like basic cruise control, you don't have to "turn it off". It just doesn't work until you don't enable.

Unfortunately on that particular car it was on by default and short of sabotaging the hardware it would stay on.

That's nice, but all of the older cars are still on the road.

Yeah, you're right. I haven't had a phantom brake in quite a while, and even the ones I used to have were not so aggressive as to be unsafe. Unfortunately, I suspect downvoters are conflating 'observing phantom braking has been fixed' with 'complete endorsement of everything Musk does'. Black and white thinking and whatnot.

Any car with regenerative braking is obnoxious, by federal law they aren't required to flash the brake lights when decelerating, ONLY when they're applying the physical brake and not when the car is regenerating, etc.

This is not true for the EU.

> 0,7 m/s² and ≤ 1,3 m/s² The signal may be generated. > 1,3 m/s² The signal shall be generated.


I don't know about the law, but that's not true of Teslas. Deceleration triggers brake lights. Here's one example where a guy tests it https://youtu.be/dmptMElBpjI

I agree, non-Teslas should be banned just for that reason. On top of that literally every other brand is also more dangerous to drive[0].

0: https://www.euroncap.com/en/results/tesla/model+y/46618

The conman has already moved on from FSD to humanoid robots that can currently spread their legs with the help of 2 other people holding the robot. It’ll keep the hype up for another decade or so, and by that time there’ll be another wave he’ll masterfully ride on.

Thunderf00t points to an interesting interview of karpathy where he says (truthfully) that when he joined Tesla, the cars couldn’t even follow the lines or something, yet at the same time the scam king was promising coast to coast self-driven rides. See timestamp 16:48 of the video below:


P.S. no offense to the excellent employees of Tesla who’ve achieved tremendous things, including the driver assist system. But the scammy claims of the conman in chief straddling the border of fraud and and futurism unfortunately overshadows these legitimate achievements.

IMO Tesla would be better if they focused solely on making good EVs, rather than also trying to become the leader in self-driving.

I get that they're trying to create a wider moat against their competitors, but if their self-driving software are found to have systematic failure modes in them as these internal docs seem to suggest, then that could very well do more harm to their reputation in the long run.

Elons point on cars being worth 5x as much with self driving shouldn’t be dismissed. If a car that costs $25,000 to manufacture all of a sudden can be driven 40-50 hours a week automatically, how much is that car now worth?

So I totally disagree on that end, as do most investors given teslas enormous valuation.

Elon doesn't actually believe that Tesla cars are about to quintuple in value now, any more than he did 7+ years ago when he first started saying it; that's just marketing BS.

> So I totally disagree on that end, as do most investors given teslas enormous valuation.

Yes, and most Gamestop investors think the company is going to be the next Amazon. Buying a popular meme stock doesn't make their predictions right.

And he’s admitted self driving was way more difficult than anyone realized, so his dream is gonna be delayed 5-10 years like most of his other major ones.

It’s now 2013, want to see a video of Starship taking off?

> that anyone realized

Not to be pedantic, but lots of people "realized" (or predicted) this, myself included.

He promised something, charged for it, and had little to show for. To me that's a big mistake, not little "ooopsie shit happens".

And it's not clear this will be fixed 10 years from now. Experimental research often fails in some stage between research and production.

Rocketry and self-driving are completely different domains of difficulty. Rocketry is a solved problem - the basic math works, we've done it before, the challenge now is building upon that knowledge, refining it and increasing efficiency, but the basics have been figured out and proven since we've launched the first satellites in orbit. SpaceX is successfully doing this.

Unsupervised self-driving in an unconstrained environment is a completely different game. The reason humans can drive is that they have millions of years of evolution in both low-level image processing, reasoning, etc, diverse knowledge of the world that an individual obtains during the course of their life which can be blended in various ways to make driving-related decisions, as well as a sense of self-preservation that generally forces them to err on the side of caution. Short of a major AGI breakthrough, safe self-driving on existing road infrastructure is impossible.

> Rocketry is a solved problem - the basic math works, we've done it before, the challenge now is building upon that knowledge

This is not true. The only thing that enabled the reusability of SpaceX rockets was the convexification of the soft landing problem, a novelty by Lars Blackmore.

Sure, you could argue the the model predictive control scaffolding was pre-existing, but so was linear algebra for self driving.

Countries have failed at rocketry. It's not a solved problem.

> > And he’s admitted self driving was way more difficult than anyone realized,

Sure and the guy is getting paid right now in cold hard cash, while all the synchopants keep ranting about hopes and dreams.

As Jack Nicholson said in 'the Departed' and I quote :

"What we generally do - in this country... is one guy brings the items and the other guy pays him. No tickee, no laundry!"

First the self-driving car for all Americans (the item), then the big payday for Mr.Musk, that's what happened with Mr. Ford back in the day, but what did the guy know? Nowadays you can just bluff your way into riches using the marvelous phenomenon of cult of personality and stock appreciation.

> > It’s now 2013, want to see a video of Starship taking off?

Want to see a sample of what Microsoft's ChatGPT can do?? It's amazing! But that has nothing to do with Microsoft giving Apple a run for its money in the smartphone market like Ballmer predicted back in 2012

There is no correlation between the 2 things just like there is no correlation between rockets and self driving cars, and it's laughable that synchopants think they are correlated just because thought leader and cult of personality enthusiast Elon Musk is both the founder of SpaceX and the petty individual who sued for the right to be called the founder of Tesla

My point was Elon has an incredible track record of success which follows a formula of overhype, delays, and eventual achievement if not overachievement.

You seem to have a more personal vendetta against him, I’m just looking at his insane success and envying hype/ management style

> My point was Elon has an incredible track record of success which follows a formula of overhype, delays, and eventual achievement if not overachievement.

His biggest success is the rocket industry. Let's examine top competitors - corrupt ULA which flies rockets on 50 year old engines and does cost+ contracts, his other competitor was Russia, a corrupt kleptocracy that stopped innovating when USSR fell apart. Russia would sell engines physically made in USSR to ULA. This market literally had no competition and no innovation for like 40 years.

So yes, his great achievement is that he could march finance into an industry that no-one would consider for investment. But results in other industries were not so great.

Tesla had a head start on electric, but they could not take advantage of this. Once other manufacturers woke up to the threat, it became clear Tesla has nothing on them, and car industry is fiercely competitive. His leadership was not good enough to take advantage of the head start he had and extremely high level of finance that he had.

Where is cybertruck? Other automakers are already selling electric trucks. VW and even Chinese automakers are eating Tesla's lunch.

Elon's boring company is an abject failure, as is hyperloop.

“ Tesla had a head start on electric, but they could not take advantage of this. Once other manufacturers woke up to the threat, it became clear Tesla has nothing on them, and car industry is fiercely competitive.”

Is there any other car company in the world selling electric vehicles at a profit besides Tesla? Everyone else is losing money on their EVs most are losing a lot of money. Not only is Tesla profitable but they have better margins than many companies have on their gas cars.

If you are looking to buy a car in 2023 and you are not looking at a model y or a model 3…

Tesla is the only car to drop prices in 2023, and the only car that doesn’t feel like you are paying 50% more for whatever you got 3 years ago.

Anything else at the 40-50k usd range is simply a joke compared to model y, and Tesla is simply eating the market right now.

Nah Tesla is just goosing the government subsidies hard. The ioniq5 is a great car just doesn’t qualify for inflation increase act subsidies

If that's the case, why Tesla easily beats any other opponent overseas as well? Here, in Australia, they do not get any subsidies. Model 3 and Y sales are leaving all others, including Ioniq, far behind.

Other manufacturers haven’t had time to scale up production. Also the direct you consumer model has been a real winner in preventing dealer price gouging due to shortages

If someone built a dozen companies, and like a third of them succeeded and made him one of (if not the) richest man in the world, I think you could say that that person "has an incredible track record of success"

and I don't even like this guy, but this is just dishonest from you

What was successful after Tesla and SpaceX?

Parent might be including PayPal too. Though it's not clear what Musk's involvement in their success was.

By the way, if you or I owned either one of the companies you mentioned it would be considered a great success. Your question itself is putting Musk in a league of his own. I don't care that he's an asshole. If someone asked the three of us to eject into space a car that we built, only he could comply.

I would say that PayPal was a success because of the mafia (which mostly consists of Thiel and Levchin's team by the way). I would argue that Musk actually hampered PayPal in many ways (pushing Windows as the dev environment, as one example, when the rest of the team wanted Linux).

That being said, SpaceX and Tesla are successful, even if they took government subsidies to get started. There are many more companies which took a lot of government money yet managed to return nothing except fat bonuses for their executives. Or the number of large companies which have stopped innovating now (IBM and Intel for instance).

  > What was successful after Tesla and SpaceX?
Is Tesla and SpaceX not enough of a track record? Seriously, if that is not a track record of success, what is a track record of success by your measure?

2 successful billion dollar companies is more than what most people can boast of.

> Tesla had a head start on electric, but they could not take advantage of this.

The Model Y is currently the world's best-selling car. I think that counts as Tesla taking advantage.


Agreed on some points.

Although I do think Tesla did take advantage of their head start. But it’s clear that the pandemic and issues with China and supply chains are hampering them globally.

And sure, Tesla launched successful Electric cars and woke the market up when they realised consumers actually really would switch.

But it’s not like other manufacturers couldn’t do it, they just chose not to.

The biggest thing Tesla did was actually making it happen with vehicles that were significantly differentiated and looked like the future!

In a market where companies stopped selling sedans in favour of SUVs, Teslas came along with the model 3 and then model Y, and showed people will choose electric tech over SUV form factor.

People weren’t buying ICE sedans because they were largely garbage!

  > The biggest thing Tesla did was actually making it happen with vehicles that were significantly differentiated and looked like the future!
Tesla also invested in charging infrastructure - taking care of both the chicken and the egg.

Your bias against elon is showing and it's not a good look. You're not being very rational.

I used to be a fan, then he showed up to the cave rescue operation with half a submarine, and in responce to crtiticism he called rescue divers pedoes.

So I started wondering, can he be a good leader when he can't take criticism? How does he retain best pepple if his ego is this fragile?

I am quite happy to see more people reassess him as he demonstrates capricious and callous behavioir, for exanple the Twitter aquisition whoch he first wanted then tried to get out of.

I think you will come around when he targets something yoi carw about.

No, I won't, because I don't think he's a faultless person. In fact, I think he's quite an asshole, but I can rationally separate who he is from his long list of accomplishments.

Just because you want him to be a failure or a con-artist or whatever doesn't mean he actually is.

Elon's biggest accomplishment is how he optimizes for capital growth by any means necessary and is willing to throw norms to the wind and think very big. Elon never had to be a good people person to succeed here. It is why successful companies allow brilliant jerks... they are so brilliant it makes up for their asshole nature. Elon is this, as much as it pains me to say it.

Jerk is not a synonym for bigoted asshole. Increasingly the cruelty (to his employees mainly, but hardly exclusively) seems to the point.

Why do people emphasis rationality when reasonability is much more important?

Why do people make semantic arguments disguised as philosophy instead of making a concrete point and furthering the discussion? Seems like a cop out to me. It's tempting to respond to everything online and hit enter but sometimes you don't have to if you don't have anything meaningful to say.

Cybertruck is delayed because people are still buying other high margin vehicles.

I think it would be fair to state that every single electric car on the market today would have been radically different if Tesla didn’t exist.

SpaceX has been quite successful, though it also came really close to failing.

Solar city actually failed as an independent company.

Boring company is floundering.

Twitter is having massive issues.

I think he’s done more harm to Tesla after taking it over from the original founders than been a benefit. It’s unclear but he definitely took massive risks which were unnecessary and has many serious failures such as how long their truck has slipped.

All together not a bad track record, but also not nearly as impressive as many people seem to think. I don’t want to suggest it’s luck, but many companies that might have been successful which came that close to failure simply failed. We look back on people who happened to have passed those thresholds because they go past them not necessarily because they had a better approach.

> I think he’s done more harm to Tesla after taking it over from the original founders than been a benefit.

Didn't every single product launch happen after the takeover? We probably would have never heard of the company otherwise, why would we assume a similar trajectory with different leadership? It's like saying Jobs did more harm than good after returning to Apple.


Tesla was founded in 2003, Musk invested in 2004, Roadster unveiled in 2006 and entered production in 2008. October 2008, Musk took over as CEO of the 5 year old company which then went public in 2010. The model S entered production in 2012 and they discontinued the roadster.

So Tesla went public selling a product developed under the original CEO, and he took 5 years to get the next model our even after having a working EV. That said the Model S was a hit, but it was also the original founders goal to work down to more mainstream products.

> has many serious failures such as how long their truck has slipped.

i wouldn't call the cybertruck a true failure, since it's effectively free marketing, and they don't have a real obligation to make a sale.

Tesla's massive competitive moat is their battery making capacity. I do not believe the incumbent car manufacturers are able to catch up any time soon.

It would be the chinese manufacturers of electric cars that pose the biggest threat to tesla, not the US/western incumbents.

Tesla is down to 54% of US EV sales (it’s much worse globally) and it’s been falling very quickly. So, their battery moat is basically gone, and the need for the recent price drop amid such rapidly expanding EV sales is a really bad sign.

A significant part of that is they lack of a truck option considering how popular the Ford Lightning, Rivian, etc are. Alongside that is the general perception of stagnation among car buyers, the yoke was seen as a gimmick not the refresh the model S is in serious need of.

They just keep fumbling. Consider the amount of bad press they got around the undersized breaks on the Plaid or their 1 foot rollout numbers. What could have been a real halo product did almost as much harm as it helped.

This is such a silly criticism, of course Tesla's share of the "EV Car" market is going to shrink as more companies sell EV cars. DUH! But that is a silly way to view things, Toyota was never compared on their share of the "Hybrid Car Market", because that is not a real thing, just as the "EV Car" market isn't really a thing, people buy cars and to the extent they cross shop they generally do so across drive trains.

Tesla's share of the overall Automotive market is growing and that is what matters.

The Model S is a tiny and irrelevant portion of Tesla's sales.

The Halo effect / most premium products do drive sales. This is graphics card manufacturers care about the speed crown for a product launch even if their most expensive products are a trivial number of sales.

The perception of the plaid being unsafe unconditionally extends to people viewing all their cars as unsafe irrespective of actual crash statistics etc. The same thing happened with self diving car fatalities it’s a trivial number of accidents but still impacted people’s perception of the brand.

> Tesla's share of the overall Automotive market is growing and that is what matters.

The company’s stock price is based on the assumption they can ride the wave of exponentially increasing EV sales and take a large share of the global car market. Having largely squandered that opportunity the company’s prospects are far less favorable.

The boring company was always hype… we’ve known how to use TBMs to drill tunnels for years

I mean we’ve known how to use rockets for decades, but SpaceX was a good idea.

TBMs are just slightly too expensive right now. At 1/2 the price per mile a huge number of projects suddenly become very attractive. Which then opens the door for more economies of scale, further efficiencies, and in theory a very valuable company.

There are plenty of lower cost TBM options, low enough cost for my local water company to be using a mini one to dig a 12 mile sewer near where I live

Lower cost, but not cheaper because you can’t use them for the same things. Aka a toy car costs less than a real one but isn’t a cheaper transportation option.

But you can’t use the Boring Company’s TBM to drill ‘real’ tunnels - it’s build for tiny Tesla car sized tunnels (which aren’t actually that much larger than the local sewer being bored)

Trains i.e. proper mass transit, require wide bore tunnels and cost increases with the diameter of the bore

A lot of the costs with tunnels isn’t the tunnel but things like the portals that need to dissipate pressure waves as trains enter and leave at high speed

The boring company produced a smallish bore TBM and then ‘oh look it’s cheaper than a large one’, there is no innovation there

Success is measured by hypothetical subtraction.

If you disappeared Tesla, SpaceX, Boring, Twitter and even Mr.Musk overnight, the world would look exactly the same the following morning.

Try and disappear Apple, Microsoft, Cisco, Boeing, JPMorgan, BankofAmerica, Exxon, Royal Dutch Shell overnight.

You'll have chaos and civil unrest within a week. That's where your quality of life comes from, sure it's not as exciting aas drinking the kool-aid of future Mars colonies and self driving cars and 100% domination in the EV cars sector, but at least it's real

Actually, Ukraine said they would have lost already without starlink.

Hardly, Starlink has had some impact on Ukraine but not nearly that much. They have called it “critical” but they call a great many things critical, the actual use of Starlink still isn’t that common.

The important bit for Ukrainian is Starlink is subsidized by other countries and it reduces the incentives for Russia to destroy communication infrastructure. Unfortunately because Russia is focusing less on communication infrastructure they get to focus more on other targets.

People following this closely disagree.

A few of those people do, but it’s a minority opinion.

People say same seriously crazy shit about the Ukraine war, but the ground truth is fairly pedestrian as far as recent wars go.

> If you disappeared Tesla, SpaceX, Boring, Twitter and even Mr.Musk overnight, the world would look exactly the same the following morning.

No it won't. Where would the Twitter trolls go? They'd flood the streets. If Twitter had disappeared no 1 can announce for president anymore!

> Where would the Twitter trolls go? They'd flood the streets.

Compare this argumentation with typical argumentation about violenceful computer games players which are also going to flood the streets for some different reasons but with similar unwanted outcomes.

I assume the person you're replying to wasn't being entirely serious :)

> wasn't being entirely serious :)

Thanks for understanding.

Easy there, cowboy, he’s had incredible success whipping his fans into frothing frenzies of self congratulatory adulation that would make a seesaw get dizzy. No other boss person since the mighty Steve has had this effect on the mass psyche, and Elon has arguably done a lot less. I claim teslas success is in spite of him, not because of him. Facts matter, and the SEC case was where I first realized his lack of leadership skill, and boasting about tanking his own stock price was not impressive. He’s not impressive, rather dull and low imagination if you ask me, which you didn’t, as you were busy falling over yourself to worship Elon Musk, who sounds like a cheap cologne

...this...has to be a troll right? The guy helped bring space flight out of a stagnant era into commercial viability and put electric cars into the world's popular psyche.

> My point was Elon has an incredible track record of success

Um what? He has a rocket company where he basically has absolutely no involvement at all...

> …while all the synchopants keep ranting…

We definitely now live in a world that needs this word. I am now adopting it myself. Thanks!

Do you mean “sycophants” or is “synchopants” a new word I’m not trendy enough to know yet?

Elon is paid in stock not cash.

that he regularly sells.

And is bought by people that think the company is valuable enough to give the stock said value. You can argue that those people are wrong, sure, but I don't get what point you're trying to make. He isn't taking money away from someone else or being given "underserved" rewards, when the people that are keeping the value of what he receives believes it to be of such value.

Point is that you first sign off the effing self-driving car, then you get your effing money.

That is what we used to do in America, since the days of JD Rockefeller, Henry Ford...but also Steve Jobs and Bill Gates and Larry Ellison. I mean I used Windows before Gates became richest man in the world and so did hundreds of millions of people.

Musk got to #1 by selling go karts to a handful of rich people and promises of future selfdriving and total domination of the EV sector when EVs are not even sure to be the winner tech in the decarbonization of cars

The vast majority of Americans have never even set foot in a Tesla.

You seem to condone the poker game that Wall St. has become, I don't agree with that because entrepreneurship shouldn't be about bluffing and then offloading your hand to a bigger fool. We do that in Vegas, like men where at least you have to look in the eyes the dude you are screwing over.

> The vast majority of Americans have never even set foot in a Tesla.

Not so sure about that. Model 3s seem to rapidly replacing Civics as the Uber car of choice.

How do you expect self driving cars, or any extremely hard technology, to come about then without long term risk capital via public/private investment? Just pure R&D working out of universities and gov labs? Let them perfect it and only then try to raise capital? Go back to the Rockefeller days before there was venture capital and mature public markets?

so what do you propose is to be done? you simply stop people from buying tesla stock so that the price doesn't go up and he doesn't get as much money? Seriously i'm really struggling to understand what you are proposing here.

You stop the fed from printing so much money. I mean, right now it's too late, but for the future, it would be a good policy.

> And is bought by people that think the company is valuable enough to give the stock said value.

They say a sucker is born every minute

> no correlation between rockets and self driving cars,

This is actually false. Automation of vehicle controls has a surprisingly large overlap between air/space and land.

The main difference is the amount of traffic and lack of vibration dampening.

If his dream involves people dying using FSD along the way, it's a nightmare.

"It’s now 2013, want to see a video of Starship taking off?" Can they even build a strong enough launch pad to handle starship?

That is awesome!

Personally I believe it, I'm not sure about the year, but someone is going to make a fully autonomous car in the next few years.

If you look at just CLIP, Dino, GPT-4, given the current pace of progress, it's pretty obvious that it's going to happen.

So it happens, then people realize they can earn money with it, then tesla realizes they can sell it now for 4 times more. In the end, you're not going to earn 4 times more because Tesla or whoever makes the car will take that into account and skim off that top.

As with everything, competitors will prevent a single offering from being priced too high.

I personally don't think Tesla will ever develop the FSD architecture that ends up used by everyone else.

Whilst I agree with Elon that _technically_ FSD could be done only with cameras, because that's how humans do it, humans only drive "fine" and not "perfectly". We need cameras, LIDAR and whatever other sensors we can fuse in.

The software that runs the thing will not be developed by a car company, because car companies (yes, even Tesla) are absolute dogshit at software. All of them. Every single one. They can't even make a proper entertainment system that isn't laggy, let alone all of the legwork & research required to piece together enough models in the right way for FSD that meets the lofty standards of the public (ie a perfect 0 fatalities, vs angry evolution-driven hormonal monkeys in multi-tonne hunks of steel powered by explosions whose fatality count grows every day at 3700 deaths/day, 8th leading cause of death globally).

>Yes, and most Gamestop investors think the company is going to be the next Amazon. Buying a popular meme stock doesn't make their predictions right.

Unrelated but GameStop should combine combine games plus what RadioShack used to be for 'us', for modern times / tech / electronics

> Buying a popular meme stock

I don’t know that I would connote “meme stock” and “Tesla”

> Elon doesn't actually believe

What's your evidence of this?

The physical capabilities of the car do not increase with the # of hours it is driven. If you work a car 5x as hard, it will require a similar increase in maintenance and an overall shorter lifetime. (Lifespan. As measured in calendar years, not miles driven)

I’m not saying you’re completely wrong on the general concept, but the relationship between value increase and drivable time per week (or whatever unit of time) is not 1:1. There’s going to be some loss in there.

Also not all cars would see drive time increases. Those used as Ubers or similar certainly will, but that won’t displace the vast majority of car owners who will continue to own a car and drive it at similar rates for the convenience of having it immediately on hand, at all times, ready to go. If it self drives too then that’s great, the claw back some personal time during a commute etc.

Good point. You're likely to get 200,000 - 300,000 miles out of a Tesla. That might take 5 years if used as a robotaxi or 20 years without, but the cost per mile and thus the value of the car will be comparable either way.

That might be different with an LFP Tesla, which have what should be 750,000 mile batteries.

Even if that reasoning were true, the valuation for Tesla doesn’t make much sense, given they’re not remotely close to a self-driving car in the way that the average customer would understand it.

They’re not delayed by a few years. The problem is intractable, unless you can limit the journey to a controlled scope.

I drive over multiple pathways and grassy lanes a week that aren’t mapped and have no markings. A self-driving car can’t arbitrarily account for something like that any time soon.

Waymo is able to operate a limited route in Phoenix because they restrict to best-case routes.

FSD is a very small component of Tesla' current stock price. Some dumbasses are certainly buying Tesla because they think FSD is right around the corner, but dumbasses tend to not have a lot of money.

For any Tesla price estimate, you've got to pick a number, how many million vehicles the company will ship in say 2025.

Then pick an Average Selling Price, profit margin, and P/E multiple.

Divide by 3 billion to get the price per share of your estimate.

Considering Tesla is on track to deliver 1.8M vehicles this year and the delivery growth trajectory it's been on, 4M is a good target goal for 2025, $45k ASP, 10% profit margin, 30 P/E . . . from that I get $180, so today's price is fair as a base case for 2025.

$400 was a nice top in 2022 and I think we'll see it before 2025 as Tesla Energy starts contributing to the bottom line, each $1B of profit from energy is worth $10 on the share price (30 P/E, 3B shares).

If you think $180 for Tesla is way too high you have to account for that in the model...either 2025 deliveries will be a lot less than 4 million, or 10% profit margin will come way down, or P/E will come way down below 30 (it's currently 51).


This is not the way I would do stock valuation. PE is the ratio of future earnings will all its growth to current earnings, discounted. In the future, some of that growth will be in the past, and the PE will be expected to come down. This being Tesla, you could imagine some new fancy product, and not value it strictly as a car company.

Edit: I see that you actually addressed the PE roll down. Missed that on the first read.

>dumbasses tend to not have a lot of money.

If wealth made people smart or gave them taste, we'd know. Dumbasses do tend to have a lot of cash, that's why they can happen to be dumbasses repeatedly.

Some dumbasses do, most don’t.

> Considering Tesla is on track to deliver 1.8M vehicles this year and the delivery growth trajectory it's been on, 4M is a good target goal for 2025, $45k ASP, 10% profit margin, 30 P/E . . . from that I get $180, so today's price is fair as a base case for 2025.

That's a lot of assumptions. Why not wait 3 years till all those projections materialize, and buy it then for $180 (if the price is still there)? Currently, if you buy at $180, you're discounting all the uncertainty and risk included in your calculations. Not to mention, even if you're right, the stock will be worth $180 in 3 years from now, which means your money will be idling for 3 years if you buy now.

This. It’s being priced as if the best possible case has already happened and can be taken for granted with 100% certainty, which if that’s the case, still means you’d be better somewhere else for the interim.

Not really, since you’re getting everything else the company is working on essentially for free at this price, including energy, FSD, the robot, the AI hardware, etc. Best case would be all of those things developing into their own revenue streams.

> If a car that costs $25,000 to manufacture all of a sudden can be driven 40-50 hours a week automatically, how much is that car now worth?

That week of driving is worth what an Uber driver’s take-home pay would be for the same week after finance (or equivalent lost interest if you buy with cash), insurance, on-road costs, charging, maintenance, depreciation, and whatever intangible value you put on not having hundreds of strangers farting in your car every week.

Multiply that by the number of weeks that you expect the car to be nice enough to do this, and that’s your return on investment.

Many people don't want to use their car as a robotic rideshare for hire. The inside is gonna get pretty not nice in a hurry.

At least when there's a human driver in a car, they can tell misbehaving people to knock it off. Not so when the car is unattended and robotic.

Less as that cost component is removed and replaced with an ROI on the cars costs.

>Elons point on cars being worth 5x as much .... all of a sudden can be driven 40-50 hours a week ...

Look at it this way, assume humans drive, or need to be driven/travel, X miles per year.

Today, assuming a car drives Y miles per year, the market needs X/Y cars.

Given your argument, with self driving, a car runs nY miles per year, the demand for car will drop to X/nY, basically 1/n of before or maybe a little more, since after all the humans won't suddenly need to drive/travel nX miles per year.

So, if a car were to become worth 5x more, but we need 1/5x cars, the total market would remain the same.

Self driving will increase demand, not decrease, as it will displace other less popular modes of transit. Why call an Uber or taxi after a night at the bar when you can just have your EV drive you home?

Why make a medium-distance flight, when you can just sleep in the back of your EV as it cruises down the highway overnight? (This will pare well with lay flat front seats what would allow turning the car into something like one of those airline sleep pods).

Self driving will increase demand, not decrease, as it will displace other less popular modes of transit

Are you sure? How many people right now own a car that they use mostly a few times per week? Car ownership right now is the only way to guarantee personal long-range transport. But if self-driving cars can replace taxis and even some form of public transport, car usage might go up but it's not clear to me that car ownership will go up along with it.

> How many people right now own a car that they use mostly a few times per week?

A lot fewer than the average SF/NYC bubbler probably thinks. It is the only viable means of transportation for probably 90% of americans.

If cars drive themselves, you can let them out as Ubers when you don’t need to drive it. That could definitely mean less cars if they are priced aggressively. Peak is still a problem, but public transit can focus on peak rather than non peak (assuming Robo taxis are price efficient).

Even if we need 1/5 cars (and EVs do last longer, so without self driving, this will happen anyways), as long as the lost cars come from your competitors, it’s fine.

> If cars drive themselves, you can let them out as Ubers when you don’t need to drive it.

This does not require self driving btw, but a more important question is: would you really? Would you let strangers in your private car? Would you want to clean it afterwards? Would you accept damages other people do to you car? Would it still be your car if strangers use it far more than you do? And at this point we have arrived at car sharing, which already exists today...

> as long as the lost cars come from your competitors, it’s fine.

It’s a rather arrogant, and unrealistic assumption, (on Tesla’s part if they are making it) that they will capture the entire market. It’s not going to happen.

> If a car that costs $25,000 to manufacture all of a sudden can be driven 40-50 hours a week automatically, how much is that car now worth?

Cars depreciate based on distance driven, not on age. A car will need to be serviced every X miles and replaced every Y miles, and these numbers are independent of whether the car is driven one hour a week or forty. In no universe does driving more hours per week increase the value of a car.

You clearly don't drive in a location that heavily salts roads! Here, a 15 year old car with 50,000 km is only marginally better than a 15 year old car with 200,000km - and it's mostly because the lower mileage car has probably been driven less in the winter.

Ultra-high mileage vehicles driven over short timeframes where rust doesn’t get a chance to set would probably get into wear-and-tear of structural welds. Something like how airframes are rated for x number of decompression cycles.

First Tesla would have to make a car with plausible hardware for self driving. Not decade+ old cameras trying to do everything. Can't even reliably wipe the windshield at the right time. The idea of robotaxis is just a fantasy.

I'm not whoring my car out to Uber to let people make a mess in my car when I'm not driving it, sorry. So to me this means nothing. I'll still be in it as much as I would otherwise. Maybe a few less trips driving my wife to errands (as she doesn't feel comfortable driving, despite having a license), but otherwise it shouldn't make much difference.

The only thing I can think of is maybe it would make some trips not need my presence at all. Like I could see placing a Wal-mart grocery order, and sending my car to go to Wal-mart without me in it, they load it up, and then it comes back, and I can unload it.

Which is nice, but I'm not paying 5x the already expensive car prices in order to save a few minutes here and there.

Why does self driving impact this at all?

You can already lend out your car for other people to drive, and there are numerous companies that do exactly that as their entire business model.

So this is a serious question, in what way does self driving cars mean your personal car is going to be used more?

The only people for whom self-driving cars represent an actual real world monetary win are people and companies that transport people and things. So public transport, trucking companies, uber, etc - for them self driving would mean no longer having to pay for drivers.

For every other group self driving has no financial impact on the value of the car (except possibly tremendously devaluing any existing cars where a component of their value is current generation "self driving").

Cars aren't suddenly going to cost more because of that...

Not a much more than I'm already paying for it, because the technology will become ubiquitous and the cost should drop appropriately.

I mean it would be worth 10x as much if its also made you breakfast but it don't and it won't.

For a brief moment, that car becomes worth a lot more. But as other manufacturers put FSD-capable vehicles on the road, values might in fact fall as overall capacity skyrockets. What is certain is that the cost of a ride will drop.

Five+ years ago when more of us believed that false promise that it was happening any day now it made more sense to have discussions about these kinds of hypothetical situations.

Is there a big market for $125,000 cars? Who needs a car 40 hours a week?


How many cars does Uber own? I thought they offloaded all of that pesky "ownership" stuff to their custome.. I mean drivers / coworkers / employees / collaborating business owners / freelancers.

The point being they could eliminate the human drivers and take the whole revenue instead of sharing it.

From ridesharing company PoV right now they are only taking 30%-50% of the revenue, and they have gaps in meeting demand due to resourcing challenges.

The business model is pretty solid if self driving taxis work even if its only in urban and suburbs

Except that currently all the drivers also manage the inventory of physical cars, and all maintenance, and all legal liability. The parent's question is a good one. Owning and operating a fleet of cars is the core competency of a rental agency, not a ride-hailing app.

> and all legal liability.

People really do not understand that the Companies like Uber, Airbnb and Deliveroo make a huge amount of money from offloading liability - it's suddenly not Deliveroo's fault that the guy delivering your pizza in London does this on illegal, unroadworthy e-bikewith no insurance, and in case of a crash he is liable. Airbnb routinely rents out apartments in blocks that have this forbidden in the leave / rental agreement.

If the human is gone, then liability offloading is impossible. You have noone to blame. That liability off-load is worth more to deliveroo than is the pittance they pay him.

Self-driving won't make a difference to this structure though ?

The pitch Tesla (if you believe them) keeps making is we could rent out our cars when we don't use it to the rental networks. We still own the car ( and the liability ) and unlock RoI.

Just a ride sharing app cannot reach the scale and revenue to justify the investments they have attracted. It is not realistic to raise $25+ Billion[1] just to be a pure SaaS startup.

Both Uber and Lyft[2] run rental businesses already, the model works where you can rent a car on a weekly basis just to drive on their apps, they cover maintenance and other services.

Uber will have to find a way to maintain their growth, and Lyft probably will be acquired by a rental network. The result is the same vertical integration.

[1] Uber's equity investment so far. SaaS startups can be valued lot more even $100s of billions, but pure tech companies do not raise that kind of cash. Stripe for example despite its delayed IPOs and ESOP issues has only raised $8B.

[2] They own some of the inventory and rest are from partners I believe.

> The point being they could eliminate the human drivers and take the whole revenue instead of sharing it.

Why would then Tesla even sell such cars if they can just compete and outcompete Uber?

Musk has already stated that Tesla will start a ride sharing service that FSD Tesla owners can participate in. It will also supposedly be exclusive (not allowed to do ride sharing for other services using FSD).

Google self driving car division has an agreement with Uber now.

I don’t know about Uber but Lyft has a car rental service. Drivers can rent cars from Lyft to drive with Lyft. If the infrastructure is already there, going from “driver turns on an app to accept/reject a ride” to “we accept that automatically” is quite simple.

I don’t believe Lyft owns the cars, they just partner with the traditional car rental agencies. Uber also has a car rental service.

Which Waymo and possibly Cruise are on track to make cars for. Tesla though, nope. Their approach is fundamentally unsound and unworkable (just like Uber's internal effort was).

5 people.

It’s not worth 5x!

Cars driven commercially like that are worth less money.

Take a hybrid Toyota Camry that’s driven as Taxi almost 25 hours a day.

Are they worth more? Do they command a premium?

No, they do not. And at the end of they short useful lives their value goes to zero.

You could argue that self driving cars don’t need drivers, and so can be driven 24/7 at less cost of labour.

Even that eventuates, and we get a world where car ownership is significantly on the decline, it just means they sell less cars.

As we’ve seen with most technology commodity products, price tends to be forced downwards.

So maybe Elon is right, and a self driving car is worth 5x. But it’s not obvious to me, and I’m wary of people like Elon talking up his own book. He is a proven loser and fantasist, but also a very successful billionaire. :-/

Elon Musk is a modern day P. T. Barnum and Tesla will never have full self driving cars on the market.

It's not just the software. They are trying to make due without the kind of sensors that are required to properly sense the immediate environment (LIDAR)

How expensive is LIDAR even? My xiaomi vacuum has LIDAR.

They were considerably more expensive the first year Tesla FSD was due to be ready by the end of the year. After that I guess it was deemed easier to do magic with some low quality cameras than get certain people to admit they're wrong and change course.

Current ones still 10-20k, some advances and mass production will lower that further. And there are very different LIDARs (frequency, resolution, range!!), the ones in Smartphones and other small devices work very different.. maybe as different as the ultrasonic sensors in your car vs medical ultrasonic (beware, bad comparison actually).

> My xiaomi vacuum has LIDAR.

I don't. Yet, somehow I operate vehicles with little to no "phantom braking or unintended acceleration."

Teslas don't have these things called "eyes" that have a few millenia of "neural network training"

More importantly, Teslas are lacking a brain.

Indeed. The difference isn't the type of sensor. This oft repeated argument that Tesla's problems are down to a lack of LIDAR doesn't add up.

Hard disagree. Unless they manage to get a computer as good as a brain, they're going to need all the sensor help they can get.

Is there evidence that any amount of "sensor help" can somehow overcome the lack of a limited "brain"?

I'm guessing your vacuum isn't doing 140 miles an hour with four humans in it with an expectation of life-conseving flawless performance for the next 10-15 years. And definitely not in the rain, at night, on rough and uncertain road.

Can your vacuum see 300 meters away and plan how to clean the east wing of your estate?

I live in an apartment so I've yet to really see the reach of this thing's lidar. It maps the room almost immediately though, then cleans by first lapping the perimeter of an area then does laps to cover the inside of the perimeter like someone mowing grass. The map is cached from the last full cleaning, so if I wanted it to do some spot cleaning I can define an arbitrary rectangle and specify how many passes I want over that spot (one is always enough, the second pass never picks up anything else). I can also control it manually but the controls have some latency from the app. I love this thing, it fills its chamber entirely with cat hair that I wouldn't even see until sending it out. It also tracks maintenance and tells me when sensors or the filter needs cleaning, or if parts like brushes need replacing (haven't needed one yet).

Mine most likely can’t see 300 meters, but it definitely knows where it is and plans it route, optimizing over time. It’s kinda neat to pick it up in the middle of vacuuming, move it to a different room, then set it back down and watch it go back to where it left off.

Does it drive you to work too?

This may surprise, but LIDAR is a very recent innovation and virtually zero percent of cars (particularly historically) are LIDAR equipped. Binocular vision has a pretty impressive track record.

Binocular vision as in human vision? Powered by the human brain? That's vastly different than binocular vision powered by Tesla's rudimentary "AI"

Also our binocular vision is movable unlike cars.

Which is how we determine depth when there is occlusion ie. we move our head.

And cars have multiple cameras simultaneously in 360° overlapping coverage. Plus you have motion parallax.

Neither of which completely solves the problem of how to determine the depth of objects in the scene.

We move our head to ascertain depth. LiDAR uses laser pulses.

And yet their self-driving still sucks compared to what a person can do.

Human eye has the dynamic range of 24 F-stops, and best cameras have like 14. That's a difference of several million. Also humans have hearing, and really advanced CPU.

Tesla can't tell apart a deadly bollard from a large plastic bag flying in the wind.

I didn’t think Tesla used binocular (stereoscope) cameras.

These comments seem to confirm: https://teslamotorsclub.com/tmc/threads/does-tesla-have-bino...

They decided to forgo LIDAR specifically though, it's not that it hasn't been out very long and they haven't adopted it yet. Other car makers have been using it for a while, just as long as Tesla could have if they truly want to be cutting edge.

It might also come as a surprise that the the lack of a LiDAR, and famously disabling the radar directly relates to Tesla’s poor self driving performance compared to the actual leaders in the area like Waymo and Cruise.

Not with electrical brains, unfortunately.

Eh, Tesla has already shown the world how to make commercially viable EVs. That's a torch the risk averse legacy auto manufacturers can take up, now that Tesla has done the derisking. Now someone needs to prove out self-driving, even at the risk of some short-term reputational harm. I don't think people realize the life-saving potential in solving this problem.

You are absolutely right, but Tesla is doing more harm than good towards making self driving commercially viable. Repeatedly making bold claims and then failing spectacularly to back them up is hurting everyone's perceptions big time. Cruise and Waymo are doing a far better job.

99% of the public have no idea about their claims or progress. Tesla doesn’t advertise and most people don’t follow the rare mentions in the news.

> I don't think people realize the life-saving potential in solving this problem.

They shouldn't be harming lives to pursue this.

Fair enough.

But if you think this line of reasoning through, cars should have never been admitted on the road in the first place.

The issue I see is lack of transparency. If the % of accidents that can be avoided is provably much higher than those caused by Tesla's self-driving tech, an informed argument could be made in favor or against.

But with Tesla withholding the information in the leak there is just FUD around the whole issue instead of facts.

no, no they shouldn't have. At least not without tesla taking full accountability and liability for every accident on autopilot.

other manufacturers are slowly rolling out more self driving like tech, AND taking on liability.

Since GP hasn't responded yet...I think you missed their point. When they said "if you think this line of reasoning through, cars should have never been admitted on the road in the first place," I think they're referring to all cars. When Ford introduced the Model T, it killed plenty of people. Ford was, in your words, "harming lives to pursue this." In the process it revolutionized transportation.

> I don't think people realize the life-saving potential in solving this problem.

People do realize it. Especially people who specialize in building safety critical systems.

And they are all aware that Tesla's doesn't do anything to build actually safety-critical systems. They just treat it as building a webapp for sharing photos of cats.

In a previous job, I was hard real time systems engineer for a major manufacturer of agricultural and construction equipment (stuff that weighed tens of thousands of pounds and can kill you in an instant). Can you elaborate in greater detail what we all think?

If the term safety critical system doesn’t ring a bell to you, well most likely you didn’t work on safety critical system (which is different from working on heavy machines).

It’s a whole engineering discipline. Just Google safety critical systems and safety engineering.

In terms of automotive, ISO 26262. Tesla’s chaotic release process and hot fixes are definitive proof that they don’t do safety life cycle and change management in compliance with it.

Note the article says the failure reports are about Autopilot, not FSD which is much more advanced (but still in beta, of course).

He has to trick the investors with something.

They already have a pretty wide moat in terms of gross margin per vehicle produced. I'd rather they focus on driving down costs - $25000 is already too much for most people.

They just have to drop the ML based self driving (since it is a nonsense...) and take the easier path.

some of tesla's models are being outclassed now that other brands are hitting their stride.

The Model X was ranked 16th by car and driver which is pathetic (the impressive BMW iX was #1)

>IMO Tesla would be better if they focused solely on making good EVs, rather than also trying to become the leader in self-driving.

I don’t think they are trying to become the leader in self-driving. I think that, too, is smoke and mirrors.

I think Elon was thinking that was the way to take cars off the road. own many cars with low utilization vs share fewer cars with high utilization.

Personally, I'm surprised that the "Summon" feature wasn't matched by a corresponding "Banish" feature (go away and park yourself)

Seems like this might be a much harder problem? At least it is when I try to find a parking place.

It would be nice if there were some "autonomous valet" system where cars could drive into a lot, get an electronic "ticket" at the gate, drive to the disembark location, then autopark in the location on the ticket.

valet banish, where nobody drives your car.

In NYC for stops of a few hours or less it would often be cheaper in gas $ to have the car keep driving, circling the local neighborhood, instead of paying for parking.

Applications are open for YC Summer 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact