Automation Thread: Imagine johnny 5 squashing your job, forever

The Something Awful Forums > Discussion > Debate & Discussion > Automation Thread: Imagine johnny 5 squashing your job, forever

«‹›3 »

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Doctor Malaver posted:

Didn't some people in this thread argue that autonomous cars are still decades away? Because rain and snow, and because proof of concept is far from actual product on the market? And the thread is only a year old!

Not sure, maybe it was another SA thread and more than a year ago but in any case this poo poo is coming fast.

It's been mostly in the unicorn thread, but yes within the last couple years there were goons insisting (and I think some still insist) that full self-driving is decades away.

# ¿ Dec 5, 2017 11:39

Adbot: ADBOT LOVES YOU

# ¿ May 9, 2024 13:21

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Yes, and we're finally close-ish to having them.

# ¿ Dec 5, 2017 17:24

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Slavvy posted:

It's on the OP.

It's probably the most obvious example of a near-future technology annihilating large numbers of jobs, and the self-driving car thread is dead now.

Slavvy posted:

Ok I see. No the problem is you've missed the uncounted times when goons get bizarrely angry about, and invested in arguing about, self driving cars. To the point where it just results in massive idiotic debates about one of the less significant automation technologies as far as impact on society goes. Which is why nobody wants to talk about them.

lol, "people get into huge debates about X because nobody wants to talk about X!"

And if/when they get rolled out in the near future they'll have a huge impact on society. So yeah wrong again.

# ¿ Dec 5, 2017 23:02

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Upgrading public transportation is itself politically infeasible though. First you'd need to redo zoning all over the place to permit higher population densities and street design that balances modes rather than favoring cars all the time. Good luck doing those things without local residents (especially in the suburbs) screaming bloody murder. Then you have to somehow convince people to invest tons of money into a system that currently only a very small % of people actually use.

Okay okay, those things aren't completely impossible, but they're drat hard to make progress on even in ostensibly liberal areas. A large part of the left is basically convinced that pretty much any kind of development is bad.

# ¿ Dec 6, 2017 14:06

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Yes, I'm very aware of ST3. That's why I said it wasn't impossible, just drat hard. Just look at the car tab backlash ST3 generated in a liberal state.

I guess I should've said, "getting actually good transit (i.e. separated grade, frequent service, wide range) is politically infeasible in most of the country".

edit: for example, Portland is comparably liberal to Seattle but seems to be largely resting on its laurels now as far as transit goes.

Cicero fucked around with this message at 15:10 on Dec 6, 2017

# ¿ Dec 6, 2017 14:50

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

The problem isn't really displacing jobs, which is to the long-term betterment of society, it's what happens to the people who got displaced afterwards.

# ¿ Dec 7, 2017 14:49

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

F_Shit_Fitzgerald posted:

If automation/A.I is a hammer, then Silicon Valley tends to see everything as a nail. There's nothing these days that they don't seem to want to automate or run with robots. Nothing inherently wrong with that, but it seems to me that the zeal for technology like autonomous cars often outpaces serious efforts to identify and combat the consequences of that technology.

I keep seeing people bring this up and it's like the dumbest thing. Making companies work more efficiently/productively is how markets are supposed to work. And not even in the "negative but inevitable side effect" sense, this is a good thing. I mean even a communist government would still desire to automate the poo poo out of everything it could.

To the extent that automation constitutes a social problem, that's the government's problem to handle. Which it should, and we should be pushing for more effort on that, but what people need to deal with a changing economy, like retraining or moving stipends or unemployment benefits, it mostly doesn't make any sense for tech companies to be providing.

# ¿ Dec 10, 2017 01:30

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

GEMorris posted:

The autonomous car supporters in this thread seem to be the ones taking the most statements out of context, posting in bad faith, making passive aggressive allusions and generally acting like condescending jerks.

Just my observations from reading the last 10 pages.

This place is chock full of idiots driven by ideology who will laugh at and mock any possible technological advances while utterly ignoring any time they were wrong by either moving the goalposts or just ghosting, and it's going to happen again with self-driving cars. Once they're actually out and available to buy or take as a taxi, suddenly every one of those morons is going to be like, "well I said they'd come around eventually DUH" or "yeah but you can't self-drive while offroading in the Himalayas" or some other such bullshit. I can already see it now:

"Self-driving cars are bullshit, it's just a limited trial in one city!"

"Self-driving cars are bullshit, they're only in sunny climes!"

"Self-driving cars are bullshit, they're only in major cities!"

"Self-driving cars are bullshit, they don't work in rural areas!"

"Self-driving cars are bullshit, they're not in Myanmar yet!"

"Self-driving cars have been around for over a decade now, what has the tech industry done recently??"

Like we have here in D&D both posters arguing that self-driving cars are decades away, and also posters arguing that we've had "self-driving cars" for decades.

Cicero fucked around with this message at 16:25 on Dec 10, 2017

# ¿ Dec 10, 2017 16:12

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Maluco Marinero posted:

No, tech companies just do their best to dodge taxes and undermine any attempts to regulate business practices that are against the society�s best interests, and since said companies do their damndest to capture regulators who are ostensibly meant to be tackling these social issues, that makes them at least a little responsible for where society is headed.

It's a perfectly fine critique of corporations (although even then the government could crack down on those things if there was the political will), but you could say the same thing about telecom corporations or oil corporations or finance corporations or just about any corporation. It's not specific to tech.

F_Shit_Fitzgerald posted:

True. I don't expect empty promises that nothing will ever go wrong, or for Silicon Valley to try to figure out all the possible ways that technology can go wrong. That's an unreasonably large request to make of them.

At the same time, though, I think Silicon Valley has been moving so fast with getting autonomous cars into everyday use that they've glossed over the consequences of that technology (and others). Elon Musk (I think it was him) saying that people who are concerned about self-driving cars "want people to die" is an especially obnoxious and irritating example. Rather than explaining the ways in which computer-driven cars could help society, or allaying people's fears and concerns, Musk chose to just build up a straw man so that he could knock it down by accusing people of taking pleasure in others' deaths.

It's hyperbole to say that they "want people to die", but "are okay with people continuing to die" seems to be accurate. Traffic crashes are one of the top causes of death (might be #1 cause) of kids and young adults.

Yeah it'd be great if we could also attack that problem with transit and whatnot too, but you're always going to have some driving.

quote:

Am I saying that Silicon Valley should halt all research and production of autonomous cars, or that R & D should be subject to democratic debate? Of course not. They should absolutely continue to innovate. But the Valley also needs to get out of its tech bubble. Self-driving cars aren't necessarily the self-evident cure-all that they've made it out to be.

What the hell kind of drivel is this? Like obviously tech companies are focused on developing technology, just like any [industry] company is focused on [industry]. And it's fine to expect them to be concerned with hacking and safety, at least as far as safety goes both SV tech companies and car manufacturers seem to be rigorously testing the poo poo out of their cars. But the indirect impacts on society from advanced technology, that's for the rest of society to decide how to deal with.

Cicero fucked around with this message at 16:30 on Dec 10, 2017

# ¿ Dec 10, 2017 16:18

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

F_Shit_Fitzgerald posted:

I think it's hype to say people are okay with others dying.

It's kind of like listening to conservatives talk about police killing black people. Yeah it's not an explicitly desired outcome (for most, anyway), but they're kind of indifferent to it because of other aspects of the issue.

quote:

I really dislike how autonomous driving has been cast as the only possible savior for those who die on the road. I'm just not sure where we got the idea that the problem was so unsolvable that we had to turn to computers to "save" us. I know you're probably going to retort that humans are awful, hopeless drivers and will never get better. I certainly see my share of awful driving on the road every time I go somewhere. But I'm not convinced that there aren't other things we could try before we take this leap.

Wow, a lot to unpack here. First, there's a lot of variance here that you're glossing over. I'm a major urbanist who would love to see serious support for walking, biking, and transit in America, but I also recognize that politically that is hard as hell, and even if we all changed our minds tomorrow changing infrastructure and culture takes a long time. And heck, self-driving cars will likely make walking and biking safer too. The year before I moved out of the country, I got hit twice on my bike by cars, both situations that a self-driving car would've avoided very easily, because they weren't some crazy construction situation or extreme weather incident, just yet another driver not paying attention.

Second, even with tons of support for the above, you're going to have some driving. Just look at...the world. Even places with great transit (I live in Munich now and it's really fantastic) still have plenty of cars around.

Third, why do we have to block on "trying other things" before taking a leap that will almost certainly radically reduce the numbers of lives lost? Over thirty thousand people die in the US every year in traffic crashes, how many lives do we have to continue losing while trying other things? What's the point even when there are also other advantages to self-driving cars anyway?

quote:

What I mean is that what looks to be obvious in Silicon Valley may look differently in the "real" world. The Juicero is a textbook example of something that probably seemed like an ingenious bit of innovation in the Valley, but looked like an overpriced, overengineered piece of crap to the rest of us. It's very easy to get into bubbles of people who think exactly like you and lose perspective.

Jesus Christ. Juicero was an example of something that looked plenty stupid to plenty of people even in SV. Just because some investors bought into something dumb doesn't mean invalidate the whole industry. Although at the same time, sometimes something looks kind of stupid but turns out to be useful or at least able to generate a lot of money. People laughed at how much Facebook paid for Instagram, and now it looks like a bargain.

You're not wrong about bubbles, but I don't see how this applies to obviously very useful technology. Which there is plenty of coming from startups too, it's just that Juicero gets headlines (because it's so intensely stupid), while something useful but boring like FarmLogs doesn't.

# ¿ Dec 10, 2017 18:37

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

ElCondemn posted:

The latest I�ve read is that someone is trying to turn the tech into a home chef.

You mean the one where it literally just uses pre-recorded motions from a human chef? Seems like it'd be brittle.

# ¿ Dec 11, 2017 00:45

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Cockmaster posted:

I'm sure there are plenty of other things we could do to reduce traffic fatalities, but the important question is how they would realistically compare to removing human error from the equation altogether. Plus it generally takes years to produce real change in society on that scale

Years? Try decades. Infrastructure and culture take a long time to change. For example, sure the Netherlands is known as a biketopia now, but back in the 60's and 70's they were gung-ho for cars.

# ¿ Dec 11, 2017 13:25

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Wow, that's impressive. Seems like something like this would make it waaayyy easier to do colored manga.

# ¿ Dec 29, 2017 22:08

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Tasmantor posted:

Algorithms have been nothing but good for us so far! If only one could tell me how to dress

Algorithms work fine all the time, but "things are fine I guess" isn't newsworthy or notable so you mostly end up discussing the times when they backfire horribly.

quote:

Who writes that winner? Apple, Google or Amazon gonna get the right to tell us all how we should dress? gently caress silicon valley I don't think anyone should be able to tell you how to dress, not fashion mags, not highschool kid, not some weird group think you believe in and sure as gently caress not the people who think "move fast and break things" is a cool attitude.

So in your Bizarro version of Earth, is there some kind of law that forces people to listen to one -- and only one -- fashion algorithm?

Nobody's gonna put a gun to your head man. I'm sure plenty of people are going continue picking their clothes out the old fashioned way.

# ¿ Jan 6, 2018 15:34

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

https://www.theverge.com/circuitbreaker/2018/1/7/16854206/foldimate-laundry-folding-machine-ces-2018

Not terribly useful at this point, as it's large, expensive, only marginally faster than folding the clothes yourself, and requires that you feed in each item carefully one at a time, but maybe someday?

# ¿ Jan 8, 2018 12:34

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Teal posted:

This is genuinely comical.

Sure it's dumb now, but it's an incremental step towards something that would be actually useful.

# ¿ Jan 8, 2018 15:05

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Teal posted:

I mean, I'm not making fun of the progress in ability to mechanically handle an assorted shapes of soft, hard to label stuff, but as a product this is pretty atrocious.

For an individual yeah. I could see it being bought as a coin-op machine for use in a laundromat though.

# ¿ Jan 8, 2018 15:25

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

https://www.geekwire.com/2018/check-no-checkout-amazon-go-automated-retail-store-will-finally-open-public-monday/

quote:

The first Amazon Go grocery and convenience store will open to the public Monday in Seattle � letting any person with an Amazon account, the Amazon Go app and a willingness to give up more of their personal privacy than usual simply grab anything they want and walk out, without going through a checkout line.

Emerging from internal testing a year later than originally expected, Amazon Go is the online retail pioneer�s attempt to reinvent the physical store with the same mindset that brought one-click shopping to the internet. After shoppers check in by scanning their unique QR code, overhead cameras work with weight sensors in the shelves to precisely track which items they pick up and take with them.

When they leave, they just leave. Amazon Go�s systems automatically debit their accounts for the items they take, sending the receipt to the app.

Personally I'm stoked to see this kind of stuff, I hope it takes off everywhere. Death to lines etc.

# ¿ Jan 21, 2018 19:02

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

One thing that is troubling for the long-term, though, is the implication that you need an Amazon account in good standing to use this store. As a one-off store that's obviously not a big deal, but imagine this was found to be effective enough to where most convenience/grocery stores used it. Much like poor people being disproportionately being shut out of the banking system due to their Chex record, you'd inevitably end up with some poor people banned from these kinds of stores due to past misbehavior. There'd probably still be some stores without such a requirement, but it might be again, similar to banking, where the poor rely on overpriced check cashing joints or predatory payday loan places; the grocery stores available to them might simply be inferior.

Now obviously existing grocery stores can ban bad customers already, but having an associated account required to even enter the store makes it much easier to do so effectively. Might be similar to doing background checks to see if potential employees are criminals: rational for each business to do, but still ends up with negative societal impact when everyone does it.

Cicero fucked around with this message at 19:16 on Jan 21, 2018

# ¿ Jan 21, 2018 19:14

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Slavvy posted:

The other thing to realise is this has little to do with your convenience or your shopping experience and much more to do with having a shop that you don't have to pay employees at.

It seems pretty obviously to be about both. I mean having no lines also means you'll get more business, if I know a transaction will go faster that definitely makes me more likely to run in and buy something.

But just this by itself doesn't mean they have no employees, they still need people to restock/clean and do customer service stuff.

# ¿ Jan 21, 2018 23:21

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Yeah, I wouldn't be surprised if that was one they're working on. As far as computer vision tasks go it doesn't sound terribly difficult, although the sheer variety of valid IDs out there might complicate it a bit.

# ¿ Jan 23, 2018 12:26

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Guavanaut posted:

And if it's not sure that it's valid within a given probability then it can just flash the light for human assistance, but it only has to be slightly better at verifying an ID than a human to do away with that step.

Yeah, I'm imagining they could easily have one of those customer service videoconferencing kiosks for backup, like what I've seen at car rental places recently.

# ¿ Jan 23, 2018 12:54

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Raspberry Jam It In Me posted:

My impression is that Amazon isn't really interested in entering the retail business.

What?

They opened up some of their own retail bookstores already. They also opened up pick-up grocery stores. Now they have a convenience store. They even bought an existing major retail chain! And you think these are the business maneuvers of a company that isn't really interested in entering the retail business?

edit: Not that I disagree that they may license their stuff out to other companies. Amazon has a history of both vertical integration and being a platform for others.

Cicero fucked around with this message at 13:45 on Jan 23, 2018

# ¿ Jan 23, 2018 13:42

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

ElCondemn posted:

So is it using the tracks left by other cars to navigate? What happens when the road is completely covered? Human drivers aren�t good at that, are robo drivers doing something different? Maybe using sonic sensors or something to find the road edges?

If the roadway has been mapped in detail ahead of time this doesn't sound like it should be too hard to deal with. The car should be able to figure out exactly where it is relative to where the lines are under the snow using a combination of GPS + distance from points of reference like buildings.

# ¿ Feb 19, 2018 18:00

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Oh man is this real? Ahahahaha
https://twitter.com/EricPaulDennis/status/975891554538852352

BENGHAZI 2 posted:

I don't understand why you would blame the safety driver over a dead lady shrieks oocc

Calm down dude, no need for hysterics

# ¿ Mar 20, 2018 15:42

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

mobby_6kl posted:

Possibly. We've all seen the (other) Volvo plow into a reporter at a self-stopping demonstration, though this should have an order of magnutude better detection capabilities than that:

https://i.imgur.com/7XPr2cL.mp4

But why don't we wait for a video and telemetry before deciding what happened

That car didn't even have the person-detecting feature: https://www.citylab.com/transportation/2015/05/why-did-a-self-braking-volvo-plow-into-a-bystander/394445/

# ¿ Mar 20, 2018 15:58

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Truga posted:

https://twitter.com/caseyjohnston/status/976110416068861953

I mean that is kind of the point of having the safety driver there. These self-driving cars are in development, it's understood that they're not completely ready to drive all on their own. The person in the driver's seat isn't just there for show.

# ¿ Mar 20, 2018 16:36

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Solkanar512 posted:

No, but you certainly do see those tragedies being studied deeply to provide ways to prevent those accidents from repeating themselves. To carry out such investigations successfully requires the ability to see that technology is not infallible.

So far nobody has suggested that the technology is infallible.

# ¿ Mar 21, 2018 11:27

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Solkanar512 posted:

Do you have OOCC on ignore?

Nope. If you want to point out where he suggested SDC's are infallible that'd be great. I just skimmed through his recent posts and didn't find anything like that.

# ¿ Mar 21, 2018 12:14

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Yeah this is an enormous gently caress-up for Uber even if they're not legally liable. You shouldn't be testing on public roads if your detection is so bad that you can't spot a person with a bike slowly going in a straight line in front of your car. Like even if the classification of person/bike failed because it was a person with a bike + bags on the bike, "large object slowly moving in front of car's path" should still trigger a response.

At the very least if you're at that early of a stage, you should have two people in the car like Waymo used to (safety driver + engineer), then you probably wouldn't have the same issue of the safety driver just looking at their phone.

Cicero fucked around with this message at 12:43 on Mar 22, 2018

# ¿ Mar 22, 2018 12:38

Adbot: ADBOT LOVES YOU

# ¿ May 9, 2024 13:21

Cicero: Dec 17, 2003; Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

Teal posted:

I didn't bother to check the source or maths so feel free to dismiss this figure as bullshit but somebody calculated it and allegedly self driven cars now are statistically 40 times more lethal for pedestrians per distance traveled than driver driven cars

According to wikipedia stats + math I just did, the US has a traffic fatality rate of 1 per 88 million miles. Waymo has done 4 million miles, so at most self-driving cars would be 22 times more lethal, but I dunno how many miles all the other companies combined have driven, adding those in would lower the rate.

# ¿ Mar 22, 2018 16:53

The Something Awful Forums > Discussion > Debate & Discussion > Automation Thread: Imagine johnny 5 squashing your job, forever

«‹›3 »