Live data from Hacker News

Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

electrek.co

31–40 of 410 posts

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#31

Earlier quoted context omitted.

I actually think the stock price would go up. His detour to fascism and megalomania has chased off tons of liberal environmentalists like myself that are the target audience for electric cars. I cancelled what would have been our replacement Tesla when he was implying on Twitter there can be good reasons for sneaking into the speaker of the house’s house and hitting her husband in the head with a hammer.

Nah Morgan Stanley and Wedbush would know the gig is up, the fraud would no longer be maintainable without Elon.

Yup. The effect of the reality distortion field has weakened, but remains plenty strong enough for now.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#32
post #24
post #5

Two more years guys, just give him two more years, a few more billion, a bit more political power and I promise he'll give you your fancy self driving toy. (Repeat from 2012 ad infinitum)

Easy problem to solve. We just need to train an AI image classifier with highly annotated images of every single possible combination of road debris in every single country in the entire world. Shouldn't take longer than a month, right team? /s

Hey boss, should we set aside the grade crossing barrier recognizer for that?

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#33

> Tesla’s EV business is in decline and the stock price depends entirely on the self-driving and robot promises I don't think that's right; I think the stock price entirely depends on people seeing it a vehicle to invest in Musk. If Musk died tomorrow, but nothing else changed at Tesla, the stock price would crater.

I actually think the stock price would go up. His detour to fascism and megalomania has chased off tons of liberal environmentalists like myself that are the target audience for electric cars. I cancelled what would have been our replacement Tesla when he was implying on Twitter there can be good reasons for sneaking into the speaker of the house’s house and hitting her husband in the head with a hammer.

It may do this or that on the announcement, but if growth stops (arguably already has) and there is no hype for years, it's likely going to grind down to a normal valuation with time.

The passive investing / market cap weighted ETF complex tends to help big valuations stay big, but a company like Tesla still needs that sharp shot in the arm followed by frenzied buying occasionally in order to stay aloft (be it by traders, retail participants, shorts covering, etc).

I suppose they could replace Musk with another hype salesman, but the "hate" that Tesla gets is a big part of these upside shock cycles for the stock, because the ticker is a siren call for short sellers, who are ultimately guaranteed future buyers.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#34

I use Tesla FSD 99% of the time. I recently drove from Austin to Florida and back. The only time I had to take over was for road debris on the highway. Off the highway it’s very good about avoiding it. My guess is Tesla has not been focusing on this issue as it's not needed for robotaxi for phase one.

> The ùberphilosopher Bertrand Russell presents a particularly toxic variant of my surprise jolt in his illustration of what people in his line of business call the Problem of Induction or Problem of Inductive Knowledge (capitalized for its seriousness)—certainly the mother of all problems in life. How can we logically go from specific instances to reach general conclusions? How do we know what we know? How do we know that what we have observed from given objects and events suffices to enable us to figure out their other properties? There are traps built into any kind of knowledge gained from observation.

> Consider a turkey that is fed every day. Every single feeding will firm up the bird's belief that it is the general rule of life to be fed every day by friendly members of the human race "looking out for its best interests," as a politician would say. On the afternoon of the Wednesday before Thanksgiving, something unexpected will happen to the turkey. It will incur a revision of belief.

The Black Swan: The Impact of the Highly Improbable, Nassim Nicholas Taleb, page 40

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#35
post #13

"is a Level 2 driver assistance system that requires constant supervision by a human driver" - the reason for human supervision might have something to do with uncommon situations (debris in road being such a situation). Elon's estimates have always been off but it is irresponsible to see an obstacle up ahead and assume the computer would do something about it while the driver and passenger debate on what the said ob…

Level 2 is very dangerous because it's so good that humans are only needed in emergencies...

The safer the system, the more catastrophic the failures

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#36

I honestly don’t know if I would have seen and avoided that, it came up really fast. And based on the video it looked like a cardboard box or something not worth avoiding until it was within 2-3 seconds range.

It's easy to give you credit here—you would have seen it and avoided it. They saw it and had plenty of time to steer to the left, in the open lane, to avoid it.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#37
post #14
post #2

I thought the title meant the software crashed, but it was the car that crashed: "They didn’t make it out of California without crashing into easily avoidable road debris that badly damaged the Tesla Model Y ... In the video, you can see that the driver doesn’t have his hands on the steering wheel. The passenger spots the debris way ahead of time. There was plenty of time to react, but the driver didn’t get his hands…

The issue with ‘driver in the loop only for emergencies’ is that it’s harder for someone to be prepared to react only occasionally (but still promptly and correctly) when outlier behavior happens, not easier. For automated tools like CNC machines, there is a reason it’s just ‘emergency stop’ for instance, not ‘tap the correct thing we should do instead’. But doing an e-stop on a road at speed is a very dangerous acti…

And you don't even know until it is too late even if you are paying attention. Suppose you see the obstruction well in advance. You expect the car to change lanes. You get past the point where you would have changed lanes but there's still time. You aren't yet aware that the car is missing the obstruction. A moment later you are much closer and make the judgement that the car is making a mistake. You grab the wheel but now it's too late.

The alternative is taking control of the car all the time. Something nobody is going to practically do.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#38

It’s an accident that could also happen to an inattentive human driver. (Ask me how I know :( ) But, if you watch the car’s display panel, it looks as if the car didn’t see anything and just went full speed ahead. That’s not great. It should have slowed down and alerted the driver that something was there. I didn’t watch the complete video so maybe there’s more.

If you watch the video, the human non-drivers in the car noticed the road debris and could have reacted, but seemed not to because they were waiting for the computer to respond first.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#39
post #14
post #2

I thought the title meant the software crashed, but it was the car that crashed: "They didn’t make it out of California without crashing into easily avoidable road debris that badly damaged the Tesla Model Y ... In the video, you can see that the driver doesn’t have his hands on the steering wheel. The passenger spots the debris way ahead of time. There was plenty of time to react, but the driver didn’t get his hands…

The issue with ‘driver in the loop only for emergencies’ is that it’s harder for someone to be prepared to react only occasionally (but still promptly and correctly) when outlier behavior happens, not easier. For automated tools like CNC machines, there is a reason it’s just ‘emergency stop’ for instance, not ‘tap the correct thing we should do instead’. But doing an e-stop on a road at speed is a very dangerous acti…

And why many "real" self-driving systems that share the road with other users are often quite speed-limited, to allow them to come to a full stop relatively safely if needed, and give a human time to take over properly. (E.g. the Mercedes system will stick way below the usual speed limits, which means it's main use currently is handling heavy traffic with stop-and-go, or many of the autonomous "busses" for dense city spaces/campuses/...)

Passively monitoring a situation continuously and reacting quickly when needed is something humans are not good at. Even with airline pilots, where we do invest a lot in training and monitoring, it's a well-understood issue that having to take over in a surprising situation often leads to confusion and time needed to re-orient before they can take effective action. Which for planes is often fine, because you have the time buffer needed for that, but not always.

Re: Tesla influencers tried coast-to-coast self-driving, hit debris before 60 miles

#40

A friend of mine who loves Tesla watched this video and said "many humans would have hit that". I feel we'll be hearing a lot of that excuse.

It's their standard go-to excuse. "I would have done that too" really says more about them as drivers than anything else.
Post reply on HN