freerobby

Nah, you're exactly right. There's further evidence of this in Tesla's early disclosure that Model 3 production was tied to getting the Gigafactory online.

HN has a lot of great discussions and smart people, but it also has a handful of armchair engineers who are eager to tear down highly successful ones like Musk. Your comment offers a rational explanation for his concern and behavior, so naturally folks pushing ulterior narratives won't like it.

Don't let downvotes get under your skin. It's way harder, more interesting, and more dignified to think and respond critically and put your name on a comment like you did, than to click a little triangle anonymously every time someone challenges your assumptions. When an earnest question like yours turns grey, it says much more about the HN community than it says about you.

Your link brings me to a 404.

It was absolutely the initial implication. Musk bent over backwards to downplay that possibility of sabotage at the very first news conference. He also tweeted that it was unlikely and publicly joked about it: https://twitter.com/elonmusk/status/783567161402888192

> Thanks, this discussion is going well. I want to add that it wasn't nhtsa, it was Tesla's data which nhtsa cited without any fact checking. But since Tesla has been parroting nhtsa's citation, they have since then withdrawn their support for Tesla's data.

Yeah, I had read that previously. I understand NHTSA not wanting to be seen as a primary source of data that it did not collect, but I don't think Tesla handing over that data changes anything material here. I don't see how the source of the raw data is relevant to the report, barring a case of fraud or massive incompetence among those interpreting it.

> So, let me reiterate -- we agree that autosteer is lane keeping assist given more responsibility. If that's helpful to the driver or dangerous as it distracts him is a debate were should have with better data, which we don't have. Unfortunately, I personally don't trust Tesla giving data without any hidden unsaid caveats, given their behavior till date.

I hear you. I would distinguish though between Tesla citing data and crafting a story around it (e.g. what they did with the 1:340M miles figure), vs Tesla providing raw data upon request to a regulatory body like NHTSA or an advisory body like NTSB, and then having those bodies draw their own conclusions. So long as Tesla isn't committing fraud and the investigative methods are statistically sound, the source of the data shouldn't matter. I like this model because, by enabling Tesla to be the data source in a way that is trustworthy, we can get access to much more of it (and much more quickly) than if we had to know all possible queries in advance and force it to go through a third party. For all the recent riffraff between Tesla and NHTSA involving the recent accident, both parties have a long track record of saying they work well together.

> However, there is another topic we should talk about. Do you think improving autosteer lane detection with vision will make a car driverless capable? Because that's precisely Tesla markets their technology as, with coast to coast drive, or cross country summon, or driverless taxis. I personally don't believe a line following robot is enough for self driving. Infact, I like Google's approach more, where they drive only when the road is free of any object (3d mapped by lidar).

I agree that Tesla's neural network needs to do much more than line following, but I don't couple that with a LIDAR approach.

One reason I am so bullish on Tesla is that I think the conventional wisdom about self driving is grounded more in marketing and fundraising than in engineering. The great thing about LIDAR is that you can build a prototype-level self-driving car and put it on a real road relatively quickly. The problem with LIDAR is that its utility breaks down much faster than that of vision. If you want to drive in conditions where humans drive today -- heavy fog or rain, sleet, snow -- you can't rely on LIDAR; you have to do that with cameras and radar. Once you've achieved a system that can do that, LIDAR becomes largely redundant; in good conditions it's telling you what you already know, and in bad conditions it's telling you less than that. So I am a big fan of Tesla's philosophy here (and of Comma.ai's for the exact same reason). This isn't just projection, btw -- Nvidia has a neural network that's been outperforming humans at camera-based object detection in bad weather for over two years now. I'm not aware of anybody achieving that with LIDAR.

> Even then, I don't think that's enough either. For example, say a ball bounces on front of you. I stop expecting a child or a dog running after it. I think driving means interacting with the environment, and responding appropriately to it, and would require a general AI.

I think you're right that with self-driving cars, we will see new classes of accidents that we don't experience with human drivers. However, because humans are such terrible drivers in everyday situations, it is likely a huge net improvement to improve the general cases and have more failures around the edges. This is why examples like the beach ball, or even the Trolley Problem, don't move me. Those situations are exceedingly rare, and optimizing for them will only get us to a local maximum in the mission to save lives. The global maximum is to mitigate the consequences of e.g. texting and driving, or falling asleep at the wheel.

Hey, thanks for that citation. That was really helpful for me to understand this critique of the data. My reading is that the 40% figure still represents the switch to Autosteer, but because the first cohort contains both AEB and non-AEB data, we don't know how much of that 40% improvement to allocate to Autosteer vs how much to allocate to AEB. Makes sense.

It's too bad that NHTSA didn't present the AEB-only cohort in the report (i.e. pre-AEB vs AEB vs AEB + Autosteer instead of just pre-Autosteer vs Autosteer).

I respectfully disagree about TACC and Autosteer (or put another way: my experience with them leads me to the opposite hypothesis, and thus my burden of proof is flipped from yours). Hopefully between Elon's promised quarterly reports, NHTSA's current and future investigations, and Tesla putting more cars on the road, we'll get better data on this moving forward than we've gotten to date.

> Because that data isn't limited to just Autosteer but also includes Autobreaking and the other safety features which are more likely to have played a role in the accident reduction.

See my response to your sibling comment. AEB was available earlier than Autosteer, and it says in multiple places (both in the summary and in the graph) that the 40% figure is since the release of Autosteer. If I am misunderstanding or misreading the report, please tell me.

> There's no evidence that Autosteer has saved any lives, but there is evidence that it's taken at least 4.

That's not an accurate summary of the evidence. There are reports all over reddit, TMC, Twitter, Electrek, etc. of Autopilot preventing accidents. There are accounts of Autosteer in particular swerving when a truck drifts toward a car. If we're going to consider one-off anecdotes that go badly, we have to consider the ones that go well, too. And there's a lot of confirmation bias to fight to do that, because "Car Doesn't Crash And Erupt Into Flames" is not a headline you will ever read.

> Those stats aren't for autosteer (which is the point of contention) alone, and include AEB which no doubt saves lives.

What are you basing this on? AEB was available before Autosteer, and it says in the report that the 40% improvement was after Autosteer installation.

> What Tesla does is however take lane keeping assist and give it more responsibility of actively driving the car.

Autosteer specifically is new, but the precedent for "active control" isn't. TACC takes just as active a role in driving a car as Autosteer, and most people feel it is net helpful (regardless of manufacturer, but especially in Tesla's case because it can see things via radar that a human cannot, e.g. a second car ahead slamming on the brakes). Heck, a lot of people think even standard cruise control is net helpful! It's not clear to me why longitudinal control is okay to do actively but latitudinal control isn't; both require attention and vigilance. My belief is that people are more skeptical of active latitudinal control because it is so new. It's understandable, but it's not an intellectually satisfying reason to hold back the technology.

How about comparing Teslas before Autopilot to Teslas after Autopilot?

> ODI analyzed mileage and airbag deployment data supplied by Tesla for all MY 2014 through 2016 Model S and 2016 Model X vehicles equipped with the Autopilot Technology Package... The data show that the Tesla vehicles crash rate dropped by almost 40 percent after Autosteer installation.

https://www.scribd.com/document/337008436/NHTSA-Full-Report-...

> They have yet to prove any of the doubters wrong with regard to the model 3, sorry.

There will always be an unproven next step. The point is that in each example where we have a clear conclusion, the conclusion has been favorable for Tesla. My money is on the Model 3 continuing this trend.

> Also, the stock is not traded on fundamentals as much as you or the rest of the public believes.

I never said (nor do I think) that the stock is traded on fundamentals.

Agree with the conclusion, disagree with the reasoning.

In the end, good products and strategies win. Betting against Tesla is dumb because they've proven the doubters wrong about their products and strategies over and over again. They've stayed right on course with the ten-year plan they laid out ten years ago, and every product they've built has been loved by an overwhelming majority of its owners. Not many companies can say that, not even successful ones.

Short sellers need to tip their caps and recognize that Tesla understands what it's doing better than they do. The longer it takes them to realize this, the more money they're going to lose.

How about audible nags? How about flashing white lights on the display? How about gradually slowing the car down until you give tactical feedback proving you are in control? Tesla does a lot of things to coerce drivers to pay attention. If you check out TMC, you'll see lots of people complaining about how paternalistic and "naggy" the system is, even for those who use it properly.

I am continually surprised by how little emphasis there is on personal responsibility when this community discusses an L2 system such as Autopilot. According to both the law and the operating manual, the driver is in control at all times. Tesla warns you of this every time you turn it on. Yes, there are enough bad drivers out there that Tesla is wise to implement habit-forming nags; but drivers also need to take responsibility for how they use (and abuse) these systems. Nobody would pass the buck to cruise control for a driver who set it to 65 and then plowed into something in a moment of distraction. All due respect to the victim here -- and I feel absolutely terrible for him and his family -- but if you are paying attention and looking at the road ahead of you, there is no situation where you accelerate for three full seconds into a concrete barrier at 70MPH -- not with Autopilot, and not without it.

There are news stories like this all over because it was an ongoing saga. But the DOT's investigation showed that folks were hitting the wrong pedal:

https://www.caranddriver.com/features/its-all-your-fault-the...

Fellow Tesla owner here. Agree with everything you wrote.

Autopilot is a wonderful system when used properly. It's lousy when it's abused. Applied responsibly, it is very obviously a safety improvement to anybody who uses it. Used as a substitute for human attention, it's very easy to see how it can turn tragic (as we are discussing/witnessing).

The way Elon talks about these situations frustrates me to no end. The data are on his side, yet he continually resorts to half-truths because they are simple to present. Pretty obnoxious for a dude who is on a tear about media dishonesty. But: that is a knock against Elon, not against Autopilot. At the end of the day, I believe Autopilot does much more good than bad, and results in in more people staying alive than would without it. I'm glad we have NTSB and NHTSA to keep Tesla honest, and I'm also glad that they are more thorough and tempered than a lot of folks here on HN would like them to be.

> I'm... not so certain. Why?

In this specific case, because if a driver is watching the road in front of them, four seconds is an awful lot of time to watch yourself head toward and then accelerate into a cement barrier, all without touching the brakes. It's fine to want Autopilot to be better, but as a matter of law, it is the driver's responsibility to slam on the brakes in that situation, and move into a legal lane.

More broadly, it's because the contract of an L2 system is that a human is in control at all times. L2 systems are assistive and not autonomous. They will never disobey a driver's presets, nor override a driver's real-time inputs, even if the car thinks it is safe(r) to do so. This is a major design principle behind every L2 system, and the reason why there are no scenarios where an L2 system is considered at fault by law.

Now obviously, if there were a bug that caused an L2 system to override a user's input -- say disregarding someone hitting the brakes, or overpowering the steering wheel, that'd be a systemic failure and grounds for a recall. But we haven't seen any cases of that, and there are hardware precautions (e.g. limited torque in the Autosteer servo) to minimize that possibility.

Tesla reminds you to pay attention and keep your hands on the wheel every time you engage Autopilot. It's one of very few legal disclaimers they show you all the time on the screen, and don't give you any way of turning it off.

They're claiming that human+Autosteer+TACC is safer than a human. They're not claiming that Autopilot while you use your phone or watch a movie is safer than human driving. And they remind you of this every time you engage the feature.

Consider AEB. AEB is safety improvement on pure human driving. But it'd be hugely foolish to rely 100% on AEB to do your braking for you. It's not designed for that -- it can help you in a lot of situations, but not all of them. The same is true for Autopilot.

I don't think it's fair to compare an L2 system to an L5 system, though. My sense is that a very good ADAS (L2) system will reduce overall fatalities, while causing a small number of additional ones through human misuse of the system. This accident seems like the latter case.

People do this all the time. Look at what Toyota went through -- drivers were claiming "sudden acceleration" but in reality they were slamming on the accelerator.

https://www.caranddriver.com/features/its-all-your-fault-the...

> Will autopilot even let you floor the accelerator with a car right in front of you?

Yes. Autopilot does not override any user input. This is a major reason why L2 system failure will always be the responsibility of the driver.

> Even I don't press my brake like that when parking.

You might if you thought it was the brake and the car was accelerating.

I agree with much of what you've written. I think the main difference between our views is in who has the onus of proof to make a change in policy. You seem to be skeptical and want Tesla to prove that Autopilot is safer; whereas I see Autopilot's design intuitively as a safety improvement, and would want data to the contrary to withhold it from the market.

While we don't have a statistically significant data set to pine over, I do think we have evidence that "a driver who [falls asleep at the wheel or texts and drives] will fare better with Autopilot than without it." I use Autopilot on my way to and from work each day. It's pretty typical for AP to handle a ten minute stretch with no input from me. For any given minute of driving, there is a very high probability that I haven't given any input to the car. It follows then, that for any given minute on Autopilot during that stretch, I'd probably be fine if I were not paying attention. This stands in contrast to the alternative: if I gave no input to the car for an arbitrary minute without Autopilot, there is virtually no chance that I wouldn't cause an accident. This is where my intuition derives from.

> The main difference between AP and other safety measures is that, based on Tesla's defenses so far, there is literally no situation in which AP should be assigned blame, because it's ultimately the driver's responsibility. Whereas when an airbag fails to deploy, or just randomly blows up in a driver's face, there can be agreement that the airbag "failed".

I understand why this sounds like passing the buck, but I think this is the right policy (for the driving public's sake). People need to be responsible for how they use any L1 or L2 system, period. Again, I would liken it to Automatic Emergency Braking. There is no situation where a failure of AEB to brake would render a driver not responsible for an accident. The system adds a layer of protection, but it does not make any guarantees. I suppose if AEB or Autopilot acted in such a way that it overrode driver input -- say AEB braked hard on the highway and ignored the gas pedal input -- then you could blame AEB. And I'd say the same for Autopilot -- if it e.g. veered toward an object and overpowered a human trying to avoid it, I would blame AP for that. But as far as failures of logic, I think the driver owns that entirely until L3 systems are available.

> the question for overall safety is: does AP result in a net increase of accidents because drivers are too over-confident?

This is a good question, and I hope Tesla releases its data as promised so we can put it to rest. My own experience is that Autopilot makes it much easier for me to pay attention to my surroundings because I don't get fatigued and I'm less stressed than when I'm driving manually. I recognize that my experiences could be atypical, but having seen hundreds of similar sentiments on TMC over the years (and very few counters), it would take data to convince me otherwise.

Look, people fall asleep driving cars every day. People text and drive, eat hamburgers and drive, wear headphones and drive all the time. For whatever reason, when they do this stuff in a Tesla, critics claim that Autopilot made them do it. I don't excuse this behavior from any driver, but all things equal, a driver who does this will fare better with Autopilot than without it.

I don't see a contradiction in your questions. Yes, Autopilot appeared to save this person from injury or death. Yes, Tesla would blame him if it hadn't. And yes, they'd be justified in doing so, because it would be his own fault for falling asleep at the wheel. Autopilot, like emergency braking, can mitigate a bad situation. But should it fail, it doesn't relieve you of responsibility for getting yourself into that situation in the first place.

> This is for the better, given the number of collisions occurring with Autopilot engaged or recently engaged.

I think this is very misguided. Tesla owners aren't suicidal or some big idiot collective. They understand Autopilot's practical intricacies (from experience) better than any other group. They continue to use it because they know it massively reduces driving fatigue.

> It's too bad the Adaptive Cruise and Autonomous Braking can't be enabled without the Lane Assist.

ACC and emergency braking work just fine without Lane Assist! One pull on the stalk for ACC; two for Autopilot.

Sloppy reporting and bad journalistic incentives are real problems, but Elon's solution is uncharacteristically dumb. Even outright falsehoods can be difficult to refute, let alone subtler varieties like "lies of omission." Few folks will see them for what they are.

As for Tesla coverage, my observation is that Elon is usually correct on the merits of his frustrations, but he always responds in the most defensive and tonedeaf way possible. I think he forgets that the people he needs to convince aren't the ones attacking him, but the ones watching from the sidelines. I consider myself a huge fan of his, and the way he comports himself often makes me sad.

My biggest pet peeve is that sometimes he'll cite a half-truth rather than a real truth, even when the real truth is on his side. His defense of Autopilot is a good example -- the data showing that Autopilot reduced crashes by 40% is very compelling, but he continues to cite the 1:340M miles stat which at best has a bunch of lurking variables. I understand why he does it -- it's easier to explain and it provides a baseline against the field rather than against Tesla's pre-AP vehicles. But it's dishonest, which is a bad look, especially when you're getting all pissy at the media for its dishonesty.

The Bolt fits the EV bill if you just need a commuter car. If you want to be able to take a road trip, Tesla is the only serious option.

No manufacturer offers anything remotely competitive with Autopilot.

Hopefully both of these things change very soon.

1. Convenience of all electric (start every day w/ "full tank", no oil changes or other regular maintenance)

2a. Social mission of the company/product

2b. Carbon footprint per mile driven gets better with age of car, rather than worse like with an ICE.

3. Autopilot

4. FSD hardware ready (even if you don't believe this, it's an advertised feature and may motivate some buyers)

5. No dealing with car salespeople

That article is clickbait. Tesla contacted the owner out of fear that his car had been hacked (i.e. by someone else). That the owner stop was merely a "recommendation," warranty invalidation was the only threat they made, and "industrial espionage" was the owner's characterization of Tesla's concern, not an allegation it made against the owner. Lots of people have gone on to hack their Teslas without consequence in the years that followed.

There was another episode where Jason Hughes was denied firmware updates after rooting his car. Elon responded that no punitive action was intended and that he views white hat hacking as a gift, and they seemed to resolve it pretty quickly.

It will be interesting to see how Tesla responds to these cases when FSD is available.

edit: Here's the original post: https://teslamotorsclub.com/tmc/threads/successful-connectio...

> This evening I got a call from service center :crying: They told me Tesla USA engineers seen a tentative of hacking on my car... I explained it was me because I tried to connect the diagnosis port to get some useful data (speed, power, etc...). They told me it can be related to industrial espionage and advised me to stop investigation, to not void the warranty.... Don't know if they really seen something in the log, because I just sniffed the network. Or maybe they seen the port scanning with nmap ? Or maybe they just read this topic ? :eek:

Do you by any chance have a source on this? I believe you, this is just such a clever system that I want to read more about it.

Sort of -- in the sense that on some occasions I'll use TACC by itself without Autosteer (most notably because TACC works great in the rain, but Autosteer doesn't). My experience is that TACC alone is still a huge improvement on regular driving. Autosteer is another big improvement on top of TACC. It's a hard thing to quantify but I'd say they are similarly helpful. Obviously the effect is increased when you put them together.

Eh. The phrase "technically correct" comes to mind. Sure, it won't be formally disproven, but if Tesla doesn't deliver FSD while these cars are on the road, that will definitely (and very rightfully) be seen as an inability to deliver.

Also, I think the conversation about Lidar in this article (and perhaps among some readers here) misses a key point. Elon doesn't think merely that one can build a Level 4 car without Lidar; he thinks it's vastly preferable for a Level 4 car not to have any dependency on Lidar -- not only for cost reasons, but also for performance in degraded atmospheric conditions (rain, snow, fog, etc.). It's a difference in strategy, not merely an attempt to do more with less.

Elon stated as early as the AP1 launch event that AP1 hardware was insufficient for full autonomy. The claim of a coming cross-country summon feature does not imply that it will be retroactively made possible on every Tesla vehicle ever produced.

If there is ever a successful lawsuit brought by AP1 owners, it is much more likely to focus on the promises that were made to them but never delivered, e.g. "meeting you at the curb" from your garage on private property.