Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
How many quarters after Q1 2016 till Marissa Mayer is unemployed?
1 or fewer
2
4
Her job is guaranteed; what are you even talking about?
View Results
 
  • Post
  • Reply
Vegetable
Oct 22, 2010

Waymo supremo :c00l:

Adbot
ADBOT LOVES YOU

Tayter Swift
Nov 18, 2002

Pillbug
Cruise is in hot poo poo with the feds after injuring two pedestrians.

The Dave
Sep 9, 2003

Not Tom Cruise, got it.

dr_rat
Jun 4, 2001

The Dave posted:

Not Tom Cruise, got it.

No, no you got it right, tom cruize was just out running one day, and ran into two other pedestrians. Didn't stop for twenty meters despite their shots and screams as they were dragged a long with him.

The Dave
Sep 9, 2003

When I first read the headline my reaction was "What the hell are they doing with those Mission Impossible movies now?"

Shooting Blanks
Jun 6, 2007

Real bullets mess up how cool this thing looks.

-Blade



The Dave posted:

When I first read the headline my reaction was "What the hell are they doing with those Mission Impossible movies now?"

We could tell you, but then we'd have to kill you.

Kwyndig
Sep 23, 2006

Heeeeeey


So has there actually been a true driverless car fatality yet or is it all just Teslas in autopilot mode running down children.

Electric Wrigglies
Feb 6, 2015

Kwyndig posted:

So has there actually been a true driverless car fatality yet or is it all just Teslas in autopilot mode running down children.

My understanding is that it has been running higher than for driver driven cars but with the full expectation that over time it will improve and longer term will be quite likely improve fatality rates significantly.

It becomes a bit of a trolly problem. Do you ban driverless cars on the basis that they are demonstrably less safe than driven cars (in the same way early auto-pilots definitely caused issues on aircraft when first implemented) right now or do you wear the temporary (to arguable level of confidence) poor rates while development continues at an accelerated rate (compared to just more and more lab trials) and payoff in reduced life and environmental impact.

Oxyclean
Sep 23, 2007


Feel like you lessen environment impact with less cars, not driverless cars.

Like, maybe there's a future where people don't really own cars, and a majority make use of a shared pool of driverless cars in a taxi like system, such that you have far less cars per capita, but it's still a very low density option.

That said, it feels unfortunately more likely that will happen then convincing America and many other countries to just build less car-focused cities and better public transit.

Electric Wrigglies
Feb 6, 2015

If cars are driverless, it makes them MUCH cheaper taxis and therefore much more tenable for more and more people to rely upon ubiquitous uber or something - maybe even with ride sharing discount.

Sagacity
May 2, 2003
Hopefully my epitaph will be funnier than my custom title.

Electric Wrigglies posted:

Do you ban driverless cars on the basis that they are demonstrably less safe than driven cars
Yes.

It's not a trolly problem, what are you talking about? There is no need for these broken-rear end driverless cars to be on the road in the first place. If they pass all sorts of rigorous testing then they might be allowed.

Electric Wrigglies
Feb 6, 2015

Sagacity posted:

Yes.

It's not a trolly problem, what are you talking about? There is no need for these broken-rear end driverless cars to be on the road in the first place. If they pass all sorts of rigorous testing then they might be allowed.

Their development being much sped up (because there is only so much testing you can do before in the field is required for decent progress) will result in overall much lower fatalities and environmental impact in probably even as tight as five year timeframe but surely 10 years (when EV should also become much more ubiquitous).

We don't ban EVs because they demonstratably often have worse accident outcomes (fires, HV risks, etc), we tolerate those risks and work on improving them to get them down to levels in-line with incumbent technology.

Evil Fluffy
Jul 13, 2009

Scholars are some of the most pompous and pedantic people I've ever had the joy of meeting.

Oxyclean posted:

Feel like you lessen environment impact with less cars, not driverless cars.

Like, maybe there's a future where people don't really own cars, and a majority make use of a shared pool of driverless cars in a taxi like system

There is a better change of the US becoming a post-scarcity utopia than the average American giving up owning a car even if they wouldn't need one.

Sagacity
May 2, 2003
Hopefully my epitaph will be funnier than my custom title.

Electric Wrigglies posted:

We don't ban EVs because they demonstratably often have worse accident outcomes
We don't ban them because they don't straight up run over people.

Baronash
Feb 29, 2012

So what do you want to be called?

Electric Wrigglies posted:

My understanding is that it has been running higher than for driver driven cars but with the full expectation that over time it will improve and longer term will be quite likely improve fatality rates significantly.

It becomes a bit of a trolly problem. Do you ban driverless cars on the basis that they are demonstrably less safe than driven cars (in the same way early auto-pilots definitely caused issues on aircraft when first implemented) right now or do you wear the temporary (to arguable level of confidence) poor rates while development continues at an accelerated rate (compared to just more and more lab trials) and payoff in reduced life and environmental impact.

I think a more accurate way of framing what they are doing is turning hundreds of thousands of people into lab rats without their consent, and if that's a necessary condition for your product to improve, then it probably shouldn't be legal. Drug development would likely be far cheaper and faster if pharmaceutical companies could make test subjects out of unsuspecting patients, but we don't allow that because it's hideously immoral.

I have also never been particularly convinced by the environmental impact argument. The most immediate improvements are just efficiency gains from using electric vehicles instead of ICE cars, and that ignores the additional power requirements needed to run all of the additional equipment that makes autonomous driving possible. Much of the remaining benefit comes from technologies that are just barely on the horizon, such as interconnected vehicle networks.
Even that ignores second-order effects, such as what the impact would be on public transportation policy and funding. I'm not keen on allowing the robo jitney cab industry to flourish when it would operate in tension with walkable development and effective public transportation.

Electric Wrigglies
Feb 6, 2015

Sagacity posted:

We don't ban them because they don't straight up run over people.

yes, being burnt to death is a far better way to go, you got me with that great gotcha.


Baronash posted:

I think a more accurate way of framing what they are doing is turning hundreds of thousands of people into lab rats without their consent, and if that's a necessary condition for your product to improve, then it probably shouldn't be legal. Drug development would likely be far cheaper and faster if pharmaceutical companies could make test subjects out of unsuspecting patients, but we don't allow that because it's hideously immoral.

I have also never been particularly convinced by the environmental impact argument. The most immediate improvements are just efficiency gains from using electric vehicles instead of ICE cars, and that ignores the additional power requirements needed to run all of the additional equipment that makes autonomous driving possible. Much of the remaining benefit comes from technologies that are just barely on the horizon, such as interconnected vehicle networks.
Even that ignores second-order effects, such as what the impact would be on public transportation policy and funding. I'm not keen on allowing the robo jitney cab industry to flourish when it would operate in tension with walkable development and effective public transportation.

Anything new is essentially treating the public as lab rats. There is no such thing as a governmental policy change that is guaranteed to not have negative (often unintended and often unexpected) outcomes. Simmerly, new airliner types nearly invariably have much better lifetime safety records than the aircraft they replace. They also nearly always have had worse safety records than the aircraft type they are replacing for the first few years of their life until a few design mistakes are ironed out.

I agree the environmental benefits are at most marginal, but locked in responsible and efficient driving (cruise control versus no cruise control is a good example) will definitely have an impact as well as reducing the number of cars if ubiquitous uber allows for the reduction of "car in every garage" requirement.

Main Paineframe
Oct 27, 2010

Electric Wrigglies posted:

Their development being much sped up (because there is only so much testing you can do before in the field is required for decent progress) will result in overall much lower fatalities and environmental impact in probably even as tight as five year timeframe but surely 10 years (when EV should also become much more ubiquitous).

We don't ban EVs because they demonstratably often have worse accident outcomes (fires, HV risks, etc), we tolerate those risks and work on improving them to get them down to levels in-line with incumbent technology.

More testing isn't particularly going to improve the technology right now. The problem is that the underlying tech just isn't really ready yet. But investors are desperate to throw money at it, so there's a bunch of companies and startups all rushing to fake it by hooking a bunch of driver assistance features together into something they think can win over tech journalists.

We don't ban EVs over their accident outcomes because there's no political will to do anything about car crash fatalities in general. And that goes double for "innovative" industries, where all the CEOs have members of Congress on speed dial and will immediately call them up at the slightest hint of inconvenience to complain about regulations strangling innovation and destroying US tech leadership. So regulators are keeping EVs and "self-driving" cars at arm's length, especially during the Trump years.

There's also the fact that self-driving and EVs are important political pressure valves. If we actually want to reduce car accident fatalities or the climate impact of driving, the true solution would be to pursue policies to reduce driving. But that's very unpopular since it requires expensive public transit investments and inconvenient lifestyle changes. On the other hand, pretending technology will solve all our problems in 5-10 years is very easy and comes with barely any political cost. Selling people bullshit dreams of robotaxi networks is far easier than convincing them to take the bus.

Boot and Rally
Apr 21, 2006

8===D
Nap Ghost

Electric Wrigglies posted:

Their development being much sped up (because there is only so much testing you can do before in the field is required for decent progress) will result in overall much lower fatalities and environmental impact in probably even as tight as five year timeframe but surely 10 years (when EV should also become much more ubiquitous).

We don't ban EVs because they demonstratably often have worse accident outcomes (fires, HV risks, etc), we tolerate those risks and work on improving them to get them down to levels in-line with incumbent technology.

The testing can continue "in the field" just as well with a driver behind the wheel. One gets the added bonus of at times running the algorithm in the background to collect data about when it is wrong without having to go out to the site where it ran someone over.

The only difference is companies have to pay for a driver.

Shooting Blanks
Jun 6, 2007

Real bullets mess up how cool this thing looks.

-Blade



Main Paineframe posted:

More testing isn't particularly going to improve the technology right now. The problem is that the underlying tech just isn't really ready yet. But investors are desperate to throw money at it, so there's a bunch of companies and startups all rushing to fake it by hooking a bunch of driver assistance features together into something they think can win over tech journalists.

We don't ban EVs over their accident outcomes because there's no political will to do anything about car crash fatalities in general. And that goes double for "innovative" industries, where all the CEOs have members of Congress on speed dial and will immediately call them up at the slightest hint of inconvenience to complain about regulations strangling innovation and destroying US tech leadership. So regulators are keeping EVs and "self-driving" cars at arm's length, especially during the Trump years.

There's also the fact that self-driving and EVs are important political pressure valves. If we actually want to reduce car accident fatalities or the climate impact of driving, the true solution would be to pursue policies to reduce driving. But that's very unpopular since it requires expensive public transit investments and inconvenient lifestyle changes. On the other hand, pretending technology will solve all our problems in 5-10 years is very easy and comes with barely any political cost. Selling people bullshit dreams of robotaxi networks is far easier than convincing them to take the bus.

"Kick the can" is probably the most popular bipartisan policy solution in US history.

Tagra
Apr 7, 2006

If you gaze long into an abyss, the abyss will gaze back into you.


Oxyclean posted:

maybe there's a future where people don't really own cars, and a majority make use of a shared pool of driverless cars in a taxi like system

This is what I always hoped driverless cars would be. Basically a cheap taxi on demand whenever you want it instead of sitting on a bus that takes an extra hour because it goes to 15 stops before the one you want.

But now I realize it will never work, because without a human in the driver seat acting as chaperone, 90% of the time your driverless car would pull up and be unusable because someone used it as a toilet or trashed it or something.

Morrow
Oct 31, 2010
The fundamental issue is that the streets of San Francisco are being used as a laboratory to train and experiment with driverless vehicles. No one objects to working on them: the issue is working on them with live test dummies.

Oxyclean
Sep 23, 2007


Honestly it makes me think of the electronic scooters that have come to annoy many people because startups don't really have to have any consideration for storage or where they get left.

Like, this isn't some common good that we collectively deal with for something better, it's corporations just testing tech to gain more profits and a bigger stranglehold on poo poo. I'm not really convinced self driving will ever be as good as it needs to be - there seems like there's just too many weird edge cases and need for the AI to adapt to unpredictable situations that it's not going to make sense. I find it funny that people sometimes point out that arguments for better public transit and "15 minute cities" don't make sense for the large parts of America that are rural and spread out....but I feel like self-driving doesn't really either, because there's almost certainly much less testing being done for all the weird edge cases you might run into in those areas with the focus of self-driving being in major cities.

Kwyndig
Sep 23, 2006

Heeeeeey


Yeah your average self driving car doesn't know how to deal with large animals in the road or even small animals in the road.

Sagacity
May 2, 2003
Hopefully my epitaph will be funnier than my custom title.

Electric Wrigglies posted:

yes, being burnt to death is a far better way to go, you got me with that great gotcha.
An owner of an EV at least gets to weigh the consequences of going with an EV.

A pedestrian being mowed down by a driverless car didn't ask to participate in this particular experiment.

Evil Fluffy
Jul 13, 2009

Scholars are some of the most pompous and pedantic people I've ever had the joy of meeting.

Electric Wrigglies posted:

yes, being burnt to death is a far better way to go, you got me with that great gotcha.

And nobody has ever burned to death in an ICE vehicle.

Stexils
Jun 5, 2008

even if self driving actually is achieved in 5-10 years (a gigantic loving "if" given the industry's track record) there's no reason to think it would reduce car usage or private ownership. you could just as easily argue there would be way more cars on the road if people didnt need to actively learn the skills to operate them. or what if a ton of robotaxi firms start up and begin competing against each other because not needing drivers eliminates a major costs of owning a fleet? why would lowering the difficulty of operating a piece of technology result in fewer people using it?

self driving has always been a sideshow to distract from industry problems of companies like uber and the actual political fights of environmentalism. if the goal is reducing the number of cars on the road there are already known ways to achieve that that other countries have implemented- mass transit, urban design and policies like congestion pricing. that car companies all oppose these efforts but not self driving should tell you which one they view as an actual threat to their business.

Baronash
Feb 29, 2012

So what do you want to be called?

Electric Wrigglies posted:

Anything new is essentially treating the public as lab rats. There is no such thing as a governmental policy change that is guaranteed to not have negative (often unintended and often unexpected) outcomes. Simmerly, new airliner types nearly invariably have much better lifetime safety records than the aircraft they replace. They also nearly always have had worse safety records than the aircraft type they are replacing for the first few years of their life until a few design mistakes are ironed out.
These aren't comparable in the way you are claiming here. The companies working on self driving cars are explicitly calling their product experimental and explaining away issues as a consequence of the development process. Aircraft are required to go through considerable private testing and meet safety standards before being certified to fly with passengers. When major issues aren't caught during that process and people die, that is what gets planes grounded and CEOs hauled in front of congressional committees. The more apt comparison would be Boeing saying "we know our plane isn't safe yet, but I'm sure we could figure out the problems faster if we could start carrying passengers."

Electric Wrigglies posted:

I agree the environmental benefits are at most marginal, but locked in responsible and efficient driving (cruise control versus no cruise control is a good example) will definitely have an impact as well as reducing the number of cars if ubiquitous uber allows for the reduction of "car in every garage" requirement.
Well there is certainly an environmental cost to building cars and constructing the places needed to store them, I consider that a distant second to the impact of the number of miles driven and the infrastructure needed to support that, and there is no reason to assume that autonomous cars would decrease either of those. In your example, if people are making the same trips, then VMT is the same. In fact, it's likely higher, as the autonomous vehicle needs to travel to you from wherever it was when it was hailed. You're billing it as a transportation solution when it really solves few, if any, transportation problems.

Baronash fucked around with this message at 16:49 on Oct 25, 2023

duz
Jul 11, 2005

Come on Ilhan, lets go bag us a shitpost


Kwyndig posted:

So has there actually been a true driverless car fatality yet or is it all just Teslas in autopilot mode running down children.

There was that one in Arizona where they disabled the built in collision avoidance and blamed the safety monitor for not stopping the car for a pedestrian their system didn't warn her about.

Dirk the Average
Feb 7, 2012

"This may have been a mistake."

Electric Wrigglies posted:

Anything new is essentially treating the public as lab rats.

Depending on the industry, stuff gets tested to within an inch of its life long before it ever hits the market. My experience is with medical devices in particular, and I've run thousands of tests and written dozens of documents to qualify the testing process and execute official testing that gets provided to the FDA. And that all happens before anyone gets close to getting the device, even in the most limited of capacities. And when novel failure methods for devices are found, testing is put into place to ensure that the failure method does not occur again.

It is unconscionable to be doing field testing like this for driverless cars. If they want to gather training data, then have a person drive the drat thing around and collect the data from the drive, then let the algorithm have a crack at it at home in a simulated environment.

Mister Facetious
Apr 21, 2007

I think I died and woke up in L.A.,
I don't know how I wound up in this place...

:canada:

duz posted:

There was that one in Arizona where they disabled the built in collision avoidance and blamed the safety monitor for not stopping the car for a pedestrian their system didn't warn her about.

This thread had a fight about whether it was the pedestrian's fault, too

Main Paineframe
Oct 27, 2010

duz posted:

There was that one in Arizona where they disabled the built in collision avoidance and blamed the safety monitor for not stopping the car for a pedestrian their system didn't warn her about.

Well, the point of a safety driver is that the car isn't capable of detecting everything and can't drive as safely as a human driver yet. The safety driver is supposed to use their eyes and judgment to detect and handle situations the car can't.

That's one big issue with driver-assistance stuff that often goes unnoticed. In theory, the human driver is supposed to be paying attention at all times, and constantly prepared to see and react to the things that the car doesn't see or can't react to. But in practice, when the vehicle can mostly run without human attention, the human driver is inevitably going to let their attention wander. And that's doubly true for paid drivers, where the company is going to expect them to be spending some of that attention on other productive work. For example, the safety driver in question was also monitoring the automated systems and noting down data, even though that was a distraction from their supposed role of watching the roads.

Boris Galerkin
Dec 17, 2011

I don't understand why I can't harass people online. Seriously, somebody please explain why I shouldn't be allowed to stalk others on social media!

Sagacity posted:

An owner of an EV at least gets to weigh the consequences of going with an EV.

A pedestrian being mowed down by a driverless car didn't ask to participate in this particular experiment.

A pedestrian being mowed down by a car driven by some idiot who won’t stop staring at their phone didn’t ask to participate either but it happens every single day and when it happens it never seems to spur much discussion.

When I’m driving a car or riding a bike or walking next to the road I’m freaking terrified of other drivers, not the driverless Waymo taxi. I would 100% support a 1 strike and your license gets revoked with super hard fines for even the smallest infractions, like overspending by 2 mph. Yes I legit support fining the everloving gently caress out of someone for doing 47 on a 45 because nothing scares me on the road than other actual legit humans driving their gigantic SUVs trying to swerve to cut ahead of me to get to their destination earlier by 30 seconds.

Vegetable
Oct 22, 2010

Driverless cars are likely better drivers than all of you and are less likely to kill someone over your lifetime :shrug:

Blut
Sep 11, 2009

if someone is in the bottom 10%~ of a guillotine
The last stats I saw showed AI cars get in more frequent minor accidents, but less frequent major accidents.

I don't know if thats changed recently but even so I'd definitely rather get in multiple low speed 20km/h fender benders with AI cars rather than be t-boned at speed at an intersection by a human on their phone, or hit by a drunk human driver.

Mega Comrade
Apr 22, 2004

Listen buddy, we all got problems!
How about the best of both worlds. Decent public transport!?

Neito
Feb 18, 2009

😌Finally, an avatar the describes my love of tech❤️‍💻, my love of anime💖🎎, and why I'll never see a real girl 🙆‍♀️naked😭.

Mega Comrade posted:

How about the best of both worlds. Decent public transport!?

But that might allow the poors to come here, which we can't have.

Kwyndig
Sep 23, 2006

Heeeeeey


That'll never happen, driverless cars are pushed by the rich, they don't want the poors using the same transportation as them.

Boris Galerkin
Dec 17, 2011

I don't understand why I can't harass people online. Seriously, somebody please explain why I shouldn't be allowed to stalk others on social media!

Mega Comrade posted:

How about the best of both worlds. Decent public transport!?

Yeah, we should do that.

But at the same time we can acknowledge that it’s insanely dumb as gently caress to go all “ban driverless cars because nobody who gets hit by one consents to be a target dummy prop” while at the same time not doing anything to actually deter people from driving with their eyes glued to their phones. People aren’t going to care until they start issuing like $5000 fines for breaking traffic rules but then you’ve got issues like how it’s unjustly affecting poor people who can’t Uber everywhere.

Sagacity
May 2, 2003
Hopefully my epitaph will be funnier than my custom title.
Yes, let's add whataboutism to the discussion as well!

Adbot
ADBOT LOVES YOU

notwithoutmyanus
Mar 17, 2009

Boris Galerkin posted:

Yeah, we should do that.

But at the same time we can acknowledge that it’s insanely dumb as gently caress to go all “ban driverless cars because nobody who gets hit by one consents to be a target dummy prop” while at the same time not doing anything to actually deter people from driving with their eyes glued to their phones. People aren’t going to care until they start issuing like $5000 fines for breaking traffic rules but then you’ve got issues like how it’s unjustly affecting poor people who can’t Uber everywhere.
Public transportation is the obvious solution and it was until we stopped maintaining it as a country. Other countries make downtowns hostile to driving and they're both more efficient and safer. Unsurprisingly, electric busses exist and so do electric ground-level trolleys! The problem is already solved.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply