this post was submitted on 20 Nov 2023
114 points (96.0% liked)

Technology

58151 readers
4112 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 1 year ago
MODERATORS
top 41 comments
sorted by: hot top controversial new old
[–] [email protected] 22 points 10 months ago* (last edited 10 months ago) (1 children)

This was always going to happen. It was never realistic to expect perfect safety from the beginning, so why weren’t they ready to handle it?

That was also my concern from the beginning for Tesla’s ambitious goal. Even if we assumed they delivered as promised, met the deadline and features for full self driving. Let’s even assume they improved safety by a full order of magnitude, 1/10 the accidents, 1/10 the deaths. That would be a huge contribution to society and ought to be a resounding success. But there still be accidents, still be deaths, Tesla would still be liable. How could a public company survive this success?

We’re making decent progress on the technical part of self-driving, but the legal and popular parts are more intractable, exactly as in this situation. I’d really like to see a safety comparison between Cruise so far, and equivalent human driving

[–] [email protected] -3 points 10 months ago (1 children)

Tons of ways to structure that legally. Lobby for laws that protect makers of self driving cars, for one. Or have the occupants be ultimately responsible.

[–] [email protected] 3 points 10 months ago

I want them to have fewer reasons to rush shit out the door that isn't ready, not to have profit over safety protected.

[–] [email protected] 16 points 10 months ago (3 children)

I’ve encountered so many edge cases over my many years of driving that I seriously doubt fully autonomous cars will truly be a thing for many decades to come. Computers / AI needs to get to the point where it can intuit what to do when a situation arises that it hasn’t been explicitly trained to handle.

[–] [email protected] 3 points 10 months ago

Road specs and markings will also need to evolve to support the technology. There will eventually be roads that are approved, and others that are not, and an evolution will have to occur where roads are brought up to spec so they can join the system. It sounds silly until you realize we’ve already done all this with the road markings and lighting and grading specs and lane width specs and signage standardization and and and and everything that goes into today’s roads. Right now all the focus is on making the cars adapt to a world not designed for them but in the long run it will be a convergence of cars and road systems that will happen.

[–] [email protected] 1 points 10 months ago (1 children)

I really feel as though the easiest way would be to rip the bandaid off and allow only robo cars.

But that comes with a slew of other issues such as peoples sovereignty.

It's also depressing, because all these organizations will spend all things money rather than just build the trains that would actually solve the transportation problems.

[–] [email protected] 4 points 10 months ago* (last edited 10 months ago)

When I was reading Heinlein and Asimov and I got to their stories about rolling roads and highways I was like "ha ha, that's cute but stupid".

Now, I'm starting to wonder how far the American transport industry will really go just to avoid doing trains and buses.

[–] [email protected] 1 points 10 months ago

Just let 'em all go in a dead city, but give them features to use as facial expressions.

[–] [email protected] 13 points 10 months ago (1 children)

GM bought Elon’s crap and it screwed them.

Time for another taxpayer bailout!!!

[–] [email protected] 4 points 10 months ago

Elon Musk is rich enough to survive without a rent-free room in your head, dude

Driverless cars ultimately will have nothing to do with Elon Musk

[–] [email protected] 12 points 10 months ago* (last edited 10 months ago) (1 children)

While self driving cars seem like a good way for enterprise to bypass the cost of paying a driver, the driver’s other function isn’t just to drive the car, but to be liable for its operation.

I wonder if it’s gonna take an insurance company to push for driverless before we see any driverless cars for sale. And if insurance companies don’t want to be liable then we may never see them.

[–] [email protected] 3 points 10 months ago (1 children)

The insurance companies will go for it if the data shows driverless cars cause fewer accidents and lower claims versus human drivers, but it seems like that data will be a long time coming because right now the court of public opinion goes nuts when a driverless car hits someone while ignoring all the times that a human does the same. It makes no sense, and I hope the insurance companies can make it make sense soon.

[–] [email protected] 1 points 10 months ago

They can always put some money to back it up, if the statistics really tell them so. I guess we'll see it when it happens.

[–] [email protected] 10 points 10 months ago (2 children)

The decision came over a month after an incident in which a hit-and-run victim became pinned under a Cruise vehicle and then was dragged 20 feet to the side of the road. As a result, California Department of Motor Vehicles suspended Cruise’s permit to operate driverless cars in the state.

Why haven’t they banned all cars over the many more incidents caused by human drivers? Including incidents where pedestrians are killed deliberately, as with the cases where idiots drove straight into protest crowds?

If we’re ever going to get past the lethality of human drivers, we need to at least judge the technology by the same standards.

We’re at 80% with human drivers and they want to throw out driverless tech because it’s 92% and not 100%.

[–] [email protected] 6 points 10 months ago* (last edited 10 months ago)

There's a high probability that if that taxi was a normal human driver they also would have hit and ran like the actual offending driver in the black sedan.

At least with the robotaxi they have dashcam footage of the assailant that ran.

[–] [email protected] -2 points 10 months ago (2 children)

Why haven’t they banned all cars

Talk about throwing the baby out with the bathwater!

[–] [email protected] 2 points 10 months ago

I think we should actually ban the production of metals. Between the deaths from guns and cars, metal is a hazardous technology that needs to be proven safer before we unleash it upon our populace, much though the steel billionaires would love to profit from it while people die.

[–] [email protected] 2 points 10 months ago (1 children)

That’s what they’re doing about driverless cars though. Instead of looking at the data overall they banned all cars from a company as a result of a single event.

I’m not saying driverless cars are there or they aren’t. But let’s not pretend this is a cool headed data driven decision. This is political.

[–] [email protected] 1 points 10 months ago

It's the same set of reasons we don't have miniture nuclear reactors creating clean energy around the place

[–] [email protected] 6 points 10 months ago (1 children)

This investment is taking longer than the myopic financial outlooks that traded companies possess. But the idea of autonomous cars is not flawed.

[–] [email protected] 4 points 10 months ago (1 children)

You might think humans could at least think as long term as “it has to benefit my children in order for me to care” but the entire commercial world runs on “it has to make me as much money as possible within 3 months for me to care” and it’s just embarrassing.

[–] [email protected] 1 points 10 months ago (1 children)

Some if these companies have been at this for decades now. I don't think lack of commitment is the problem.

[–] [email protected] 3 points 10 months ago

Maybe it’s just the way the author wrote this article, but they make it sound like a 7 year investment that hasn’t matured is an epic disaster.

It’s been a tumultuous seven years since GM first announced its plan to acquire Cruise with the goal of rapidly commercializing the technology. The company has scored some significant victories in recent months, only to see most of that progress evaporate after a series of errors have exposed major problems with Cruise’s management. And now Vogt’s resignation puts GM in a tough position: continue to fully embrace self-driving cars, or cut its losses.

[–] [email protected] 5 points 10 months ago

The thing that bothers me is if you watch the first need report a human wouldn't handle the situation any better than the robotaxi did.

A human driver hit and ran and it threw the person into the robotaxi

[–] [email protected] 5 points 10 months ago (3 children)

Job #1 for self driving vehicles should be don't hit anything. Self driving should be able to have better senses, 100% attention, and better reflexes than human drivers. Until the vehicles can operate without hitting things they should never be allowed on public roads.

[–] [email protected] 5 points 10 months ago (1 children)

To be fair, the incident they're referring to was from a driver hitting a pedestrian which then knocked them into the path of the Cruise vehicle. They actually do a pretty good job of not hitting things but every incident gets amplified greatly.

[–] [email protected] 1 points 10 months ago (1 children)

Good lord is that what happened? Damn. The misdirection of attention here is astounding. I’m surprised people aren’t trying to ban the production of metal because it was involved in vehicular deaths.

[–] [email protected] 1 points 10 months ago

The issue was that the Cruise didn't detect the collision and dragged the victim along, not that it caused the accident.

[–] [email protected] 4 points 10 months ago (1 children)

Until the vehicles can operate without hitting things at a better rate than human drivers they should never be allowed on public roads.

Humans are atrocious drivers, once that mark is reached it's a massive improvement with only upward trajectory from there

[–] [email protected] 2 points 10 months ago

They too often are, but it somehow seems more accepted for a careless human to hurt someone than a machine to hurt someone.

[–] [email protected] 2 points 10 months ago (1 children)

You just articulated the “100% or nothing” standard, which totally ignores how unsafe human drivers are. Let’s say humans score an 80% on safety today (after all hundreds of thousands are killed on the roads annually). You’re saying that a technology that’s only 92% safe should not be permitted. Nope. We need that last 8% to be there and we’ll withhold the 12% improvement from the public until it is - even though that has a cost of thousands of lives.

This is a way to get nowhere and kill as many people as possible.

[–] [email protected] 2 points 10 months ago

I didn't say 100% but they are not ready to be on public roads when they hit parked fire trucks. Huge objects on the road.

[–] [email protected] 3 points 10 months ago

They fucked up by not sharing the full accident video with public officials.

[–] [email protected] 2 points 10 months ago

This is the best summary I could come up with:


The company has scored some significant victories in recent months, only to see most of that progress evaporate after a series of errors have exposed major problems with Cruise’s management.

GM CEO Mary Barra would routinely invite him to appear on earnings calls or to speak at investor conferences in a sign that the automaker was fully invested in Cruise.

Barra herself went onstage at CES in 2022 and declared that GM would sell fully autonomous vehicles, powered by Cruise’s technology, to regular people by mid-decade.

Barra reportedly told employees that GM general counsel Craig Glidden will serve as Cruise’s co-president alongside Mo Elshenawy, who will also become chief technology officer.

“We continue to believe strongly in Cruise’s mission and the potential of its transformative technology as we look to make transportation safer, cleaner and more accessible,” Barra stated in an email to employees, according to TechCrunch.

He argued that self-driving cars would lead to a dramatic drop in traffic fatalities, using the example of a young girl killed in a San Francisco intersection to bolster his argument.


The original article contains 898 words, the summary contains 178 words. Saved 80%. I'm a bot and I'm open source!

[–] [email protected] 2 points 10 months ago
[–] [email protected] 0 points 10 months ago (1 children)

Anyone who thinks that these things could or would be anything other than a colossal waste of time and resources should watch Adam Something’s recent video on them. Such a pointless way to solve inner-city transport

[–] [email protected] 2 points 10 months ago* (last edited 10 months ago) (1 children)

Okay, I watched it.

Summary of video:

  1. self driving cars are seen as a quick fix
  2. we don’t have full self driving capabilities yet - Elon overstated his cars actual abilities
  3. adding lanes has never fixed traffic - demand just fills the new capacity. Therefore self driving cars will make traffic worse.
  4. cars are generally terrible and inefficient, hurray for trains
  5. US cities were built for cars and not public transit: we should build denser cities instead of all this.

1 and 2 are beside the point and can be discarded.

3 is the core argument and is circular, essentially saying that anything that increases capacity will make traffic worse. If this seems fundamentally flawed, it’s because it is. It assumes infinite demand. You could easily apply this same logic to trains: add more frequent trains and riders will just flock to enjoy the new capacity until they are crowded again. The reality is that there is a right amount of capacity, and the question is what kinds of cars can best utilize the lane capacity we have.

4 and 5 are good points but mainly argue that we should not ONLY focus on self driving cars as a complete transportation panacea, which is true. But no one is doing that. Therefore this is a straw man argument.

The silent presumption of this entire video is that the sole, entire hope of self driving cars is to reduce urban traffic congestion. This is patently false. They also aim to improve on the abysmal safety record of human drivers, and improve fuel efficiency by taking people’s lead foot off the gas pedal, and finally to make access to a car more economical for those who don’t own one or can’t drive because of disability or age.

So basically, it’s what you’d expect from a YouTube video: some random guy leaning way too hard on a couple of limp arguments to make a sensational video that will get clicks because it has extreme claims in the title. Throw in some Elon hate and cherry picked videos of self-driving errors and the narrative is complete.

[–] [email protected] 1 points 10 months ago (1 children)

the question is what kinds of cars can best utilize the lane capacity we have.

The kind that can take 50-100 passengers instead of 1-5?

It's not about who's driving the vehicle, it's about what's a sustainable ratio of people: vehicles.

Make self-driving vehicles, by all means. Autonomy won't solve the fact that number of people in the city divided by 5 (best case scenario, but we all know it's more like 2 or 1) equals vastly more cars than there's road surface.

We have autonomous subways in Europe btw, they work very nicely and they minimize the distance between successive trains at rush hour. I'm all for driving automation but the circumstances need to make sense. Subway automation won't make up for train capacity or station capacity, for example, once a train or platform fill up they fill up, end of story.

[–] [email protected] 2 points 10 months ago

America and Europe are most definitely not 1:1 analogues, and crs will be of significant importance for the US moving forward.

Our job is to find ways to maximize their efficiency and safety, since they will exist. Driverless cars are the best method for both.

[–] [email protected] -1 points 10 months ago* (last edited 10 months ago) (1 children)

It's hilarious how many massive sectors of the economy bet billions on computers that could function like human brains, when that technology is clearly impossible right now. We might not even get close for the foreseeable future.

I'm curious whether the people working on these projects genuinely believed they could deliver, or if the whole thing was billion dollar snake oil the entire time.

[–] [email protected] 2 points 10 months ago

As long as they get a paycheck no outcome is an outcome.