this post was submitted on 11 Feb 2024
447 points (98.1% liked)
Technology
59753 readers
3029 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
This is dumb as fuck. Human drivers are literally the number one cause of preventable fatalities.
Humans are literally responsible for all preventable things in society. This take is also "dumb as fuck"
Did I says "humans" or did I say "human drivers"? We're not talking about liquidating all humans, we're talking about replacing them at the task of driving.
Totally no other spots in society where machines have replaced formerly human done tasks to make work safer, what a crazy idea that is right! /s
Yes and saying "human drivers" are often at fault when talking about smashing up motor vehicles is just as silly.
As you can also say; Hey, did you know that 100% of deaths related with driverless cars involved software?
Yes, except that we can then compare which is safer.
So far, when done gradually and responsibly, it turns out that software seems safer.
Yes, clearly that must be why they fought so hard to hide their crash data
While I agree that data should be public, them not wanting every crash to be reported on publicly probably has something to do with the fact that mobs are burning their cars down at the present moment, even though they're statistically safer than normal drivers.
Safer according to Waymo data, who has a history of and is currently hiding data.
Individual transportation is the number one cause of preventable road fatalities, human or machine doesn't matter.
Even in an ideally car-free society, you will literally never be able to get rid of taxis, deliveries, moving large furniture / household items, etc. without some form of enclosed motorized transportation (a car for the purposes of this discussion).
If we make machines that are safer than humans than yeah, it will.
That's not individual transportation. None of it. And do imagine how your city would look like if those were the only vehicles on the roads. Go to the next intersection, count cars, see how many of them would be gone, how much road surface could be converted into a tram lane, comfortable bike lanes, greenery, also, a hot dog stand.
Malaria might be less severe than the bubonic plague still doesn't mean I want to catch it.
Well the don't shift goal posts to "individual transportation" when we're talking about people thrashing a self driving car.
They didn't trash a normal individual transporter.
You can do automated taxis, deliveries and moving services, not so much.
You can still automate the driving part of moving and delivery services, which is the dangerous part.
No. Not securing loads is the dangerous part. You need a human in there anyway and with the current sorry state of driving automation best you can do is have them browse the delivery list while the car is handling a traffic jam.
There's a reason you don't see the likes of UPS or DHL get into automated cars, but venture capital moonshot tech companies promising nonsense on the one hand, as well as traditional car manufacturers with way more reasonable claims. IIRC Audi is actually leading the pack.
And it's not like UPS or DHL know nothing about vehicles, they're driving custom orders. DHL even was a manufacturer for some time.
Jesus christ, you're trying to argue that driving isn't dangerous? Ok bud, glad to see you're approaching this discussion in good faith /s
Yeah, cause they literally started from DARPA's moonshot program and take massive amounts of cutting edge machine learning to execute, not exactly DHL / UPS' strong suit given that they contracted out development of almost all of their software until very recently.
Professional drivers have a very, very low accident rate. And generally don't tend to be at fault even if they get into one. Distracted commuters are where the accidents happen, people who should not be using roads but public transportation.
What part of "DHL manufactured cars themselves" did you not understand. They know exactly what they need from their vehicles and self-driving wasn't on the list. Electric was on the list, specific range requirements were on the list, second front seat wasn't, instead you have comfortable loading heights and well thought through access to the load (that includes the missing 2nd front seat). That's the stuff that actually matters for a delivery van. Automated driving would only get into the way of the fancy manoeuvring the vans do.
Yeah, you're right, there's no point implementing any road safety standards or technology whatsoever because it would be better if we all just instantly switched to public transportation! Thank god we live in a world where it's only ever worth it to pursue the most perfect and naïve solution! All we have to do is rework our entire transportation network and tear down existing houses and force the residents to all move into villages! What a perfect solution, totally feasible in the next 10 years.
Lol, DHL didn't specify self driving because it wasn't an available option, and they don't have the technical capability to build, not because they wouldn't want it. They have an entire page on their website stating explicitly that they are closely monitoring self driving technology as it stands to have a huge impact on their business.
Like automated driving? Shit isn't working, isn't even close to working to the degree that advocates said it would ten years ago. Meanwhile, public transportation is a tried and true approach that actually fixes issues. It's vastly more energy and resource efficient and does not create socio-economic barriers to mobility.
Yards make sense and that stuff actually is in operation in many places, it's a controlled environment. You'll be hard-pressed to find a modern container terminal without autonomous vehicles.
Long haul does not make sense as that's train territory. Which of course can also drive automated which, unlike self-driving cars, actually a mature technology. Drivers are still used long-haul though because there's need to do non-driving tasks that AI can't do, automated trains are a metro thing.
Last-mile makes approximately zero sense. Parcel pickups are the right solution for standard service and for premium service AI generally won't be capable enough for decades if not centuries to come. You don't want a pharmacy to wait for life-saving medicine because someone put a traffic cone on the hood.
It is working, Waymo is operating taxi services in two cities successfully, and yeah, it's turned out to be a harder problem than initially realized, so were smartphones, now they're everywhere. You know what's a harder problem that will take longer than decades? Reorienting all of society around villages and public transportation and forcing people to move and abandon their cottages.
No one's arguing against public transportation. Stop trying to make it sound like this is a car vs. public transportation thing, when it's people trashing autonomous cars and not other cars.
And because trains don't get you to the last mile, roads do.
Again, you'd still have a delivery person for critical deliveries, they just wouldn't be driving. If someone asshole wants to stop ambulance right now they can too.
It's not that hard and no you don't need to re-do the whole country to make a massive impact. Making a random US city walkable would take like five years max if you actually set your mind to it.
Then why are you bringing up strawmen "but we need delivery vans" when I was specifically talking about individual transportation being bad? Ten cars getting replaced by a car share is good, one car getting replaced by one autonomous car is bullshit: It's not self-driving that will address the systemic issues with transportation. Just as many cars on just as impassable roads won't make them safer for pedestrians. I get being excited by a technology but it's far from a silver bullet, on its own it addresses quite literally nothing of relevance.
As do collect taxis. Do I have to repeat that five times more.
If someone wants to stop an ambulance the driver will go around them, or go out and curse their ass off until they move, or right-out push their car out of the way (because yes with blue lights on you have the right to ram, over here, the offending driver will be sent a bill for the repair of the ambulance and face criminal charges). Good luck teaching an AI to make those calls.
And you'll still have drivers, and it will still be safer if they were autonomous.
So now every technology improvement needs to solve every single systemic problem or it's not worth pursuing? That's your argument? Autonomous vehicles don't solve every transportation problem, they solve the problem of drivers regularly killing and maiming people.
Oh wow, the literal millions of road deaths every year are now "nothing of relevance".
No they're not. Not for the distances covered in many rural areas. Try and wrap your brain around the fact that not everywhere is Europe where there's millions of people packed into a postage stamp.
You literally quote the answer to that:
(Answer merged to here)
The fact that a thing most people do and some for hours daily has a large effect shouldn't be surprising
Oh wow, what good reasoning!
Let's all take up smoking cigarettes indoors all day, it's incredibly dangerous and is killing mass numbers of people on a literal daily basis, but that's fine because everyone's doing it, so the effect shouldn't be surprising, so that makes it ok and not worth addressing!
/s
That's not the argument I'm making. What I'm saying is that if you only take the raw numbers for a given event into account, and don't consider the population of the event, then you can make any event affecting a large population look like an urgent affair when it's not so urgent
Human driven cars should be replaced with automation (or even better, automated public transportation) as soon as it's viable. It's not yet, so we should not rush corporations to put their unsafe vehicles on the street. Because then the only thing you'll rush is transforming human driver fatalities into robotic driver fatalities, and you never know how worse things can get
Edit: Wording
How is this different from my cigarettes analogy? You're just arguing it's not a big deal that hundreds of people are dying on a daily basis, because a lot of people drive.
Fully agreed.
Except that it is. Waymo already has a safer per mile rating than human drivers.
Great, so anything that humans do that can potentially harm another human, should be given to the algorithms instead?
Sure, they're private for-profit blackbox algorithms, but it's obviously better then letting humans do things.
Don't worry, I'm sure once the tech oligarchs have secured just another 25% control over our daily lives, they'll start the giving back and bettering humanity parts of their business plans.
Uh yeah, once it's proven safer why wouldn't you?
Because you're scared of the word algorithm?
You seem to have an issue with wealth distribution, not autonomous vehicles.
How much could an autonomous car cost Michael, $10?
We're talking about a taxi service, not an individual's car. Waymo is not example of wealth inequality, unless the brush you use is as broad as "requires technology to run = tech bro devilry'.
Happen to know the cost of one Waymo taxi vs one taxi plus a person making a living from driving that taxi? I do know that Waymo charges more and almost all info is hidden.
In 2021 it cost Waymo ~$180k for a brand new Jaguar i-pace with all their sensors and computers outfitted. Given that 2021 i-paces started at $70k, we're looking at ~$100k for the sensors and computers necessary.
So since a taxi driver in san fran makes (according to a quick google) $48,384 a year we can assume this means they need at least 2 years out of these to break even. This is assuming it does not get set on fire from the driver who is now out of a job.
What's your point? That's self driving systems are incredibly cheap and worth it? An ROI of 2 years on first generation hardware is very good spot to be in.
I wouldn't hang my hat on that statistic until after autonomous cars make up a significant portion of cars on the road.
Fair point, but the difference is that human drivers are already at roughly their limit for how good they can drive, but self driving cars have the potential to exceed us.
It's similar to one of the biggest arguments in electric vs gas cars. Even if electric cars today are just as environmentally unfriendly as gas cars (they're not) the difference is that gas technology is super mature and there's very little improvements to be made by spending more money on it, electric battery technology on the other hand, is still in it's relative infancy and has huge potential to improve in numerous ways, but that can only happen if more people buy electric so more R&D money can be spent on it.
Ultron, is that you?