1958: "Self-Driving Issues"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

Kit.
Posts: 1079
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Kit. » Fri Feb 23, 2018 8:46 pm UTC

ucim wrote:I don't have a number for you, but it's definitely nonzero. I've been known to just get in the car and go.... second star to the right and onward 'till morning. Ok, not literally 'till morning, but enough to know that I do not need a map or GPS to avoid hitting things. I've also driven in several foreign countries without a map... including Italy and the Dominican Republic.

So, not a big percentage of your turns were non-predicted but made with the normal speed of traffic, and you don't really know your safety statistics on those?

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Fri Feb 23, 2018 9:03 pm UTC

Kit. wrote:So, not a big percentage of your turns were non-predicted but made with the normal speed of traffic, and you don't really know your safety statistics on those?
I don't even understand the question. I have never been in an accident that was due to poor navigation (location-on-a-map awareness). All of the accidents I've been involved in were due to poor control (obstacle avoidance, location-wrt-other-vehicles awareness).

None of the accidents I've been involved in would have been averted by better GPS mapping. My car's location is (probabaly) not in the GPS database of any of the vehicles that hit me.

Let me ask you this - suppose you were in an unfamiliar city or town, visited a car dealership, and wanted to take a car for a test drive. The salesman would get into the passenger seat, you would get into the driver's seat, and the salesman would suggest something like "make a right out of the driveway, go to the light, and then take a left..."

Would you be able to safely handle the vehicle with the GPS off?

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
Pfhorrest
Posts: 4888
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1958: "Self-Driving Issues"

Postby Pfhorrest » Fri Feb 23, 2018 9:23 pm UTC

ucim wrote:
Kit. wrote:What percentage of your driving happened on roads you were driving for the first time ever - without a map or of a knowledge of directions beforehand?
I don't have a number for you, but it's definitely nonzero. I've been known to just get in the car and go.... second star to the right and onward 'till morning. Ok, not literally 'till morning, but enough to know that I do not need a map or GPS to avoid hitting things. I've also driven in several foreign countries without a map... including Italy and the Dominican Republic.

I have gotten lost sometimes. But that's a navigation issue, not an obstacle avoidance issue. Do you not see the difference?

Jose

Seriously, this. Are some people that bad at driving? I once hopped in the car near midnight Saturday in California and drove to Illinois by
early Monday morning through six states I’d never set foot in and thought nothing of the “keeping the car on the road at speed” aspect of that until now.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

speising
Posts: 2272
Joined: Mon Sep 03, 2012 4:54 pm UTC
Location: wien

Re: 1958: "Self-Driving Issues"

Postby speising » Fri Feb 23, 2018 10:43 pm UTC

driving on highways is easy. an unknown city, OTOH, can get quite stressful.

User avatar
Pfhorrest
Posts: 4888
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1958: "Self-Driving Issues"

Postby Pfhorrest » Fri Feb 23, 2018 10:55 pm UTC

In my experience driving in most cities in general is stressful. Especially older ones that were not originally designed with cars in mind. Even in cities I know well, just driving around these poorly designed (even if they were designed after the fact and can't help it) city streets is just inherently going to be stressful.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Sat Feb 24, 2018 2:09 am UTC

Stress is not at issue. Yeah, in the city you are more in competition with other drivers than on the open road, but that's part of obstacle avoidance. Sure, not knowing where the turn is can add a bit of stress, but that is not a major component in inadvertent body style modifications. In the end however, avoiding the car in front of you has nothing to do with the latitude and longitude of Fifth Avenue, or how often a red car is parked on the corner.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

Kit.
Posts: 1079
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Kit. » Sat Feb 24, 2018 7:10 pm UTC

ucim wrote:
Kit. wrote:So, not a big percentage of your turns were non-predicted but made with the normal speed of traffic, and you don't really know your safety statistics on those?
I don't even understand the question. I have never been in an accident that was due to poor navigation (location-on-a-map awareness). All of the accidents I've been involved in were due to poor control (obstacle avoidance, location-wrt-other-vehicles awareness).

Is lane changing "navigation" or "control" in your terms, and why?

ucim wrote:None of the accidents I've been involved in would have been averted by better GPS mapping.

What does it have to do with "GPS mapping"?

What is that "GPS mapping" thing anyway, and why do you keep talking about it? Cars don't do "GPS mapping", they do "GPS localization". Cars do mapping with lidars, radars, cameras and (hopefully not) bumpers, they also reuse already made maps.

ucim wrote:Let me ask you this - suppose you were in an unfamiliar city or town, visited a car dealership, and wanted to take a car for a test drive. The salesman would get into the passenger seat, you would get into the driver's seat, and the salesman would suggest something like "make a right out of the driveway, go to the light, and then take a left..."

That's already navigation.

Besides, car salesmen prefer to choose safer routes. The car is still covered by their own insurance.

Pfhorrest wrote:Seriously, this. Are some people that bad at driving?

All people are bad at driving. Not everybody is able to admit it.

One flight on a modern airliner is about as dangerous as 10 km by a car.

Pfhorrest wrote:I once hopped in the car near midnight Saturday in California and drove to Illinois by early Monday morning through six states I’d never set foot in and thought nothing of the “keeping the car on the road at speed” aspect of that until now.

So what?

I once drove from St.Petersburg, Russia to Munich, Germany. Was it safe? Hell no; not in Poland at night through a thunderstorm with hail and heavy rain, with the freeway traffic jumping between 50 and 130+ km/h. Am I alive? Well, guess so.

User avatar
Pfhorrest
Posts: 4888
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1958: "Self-Driving Issues"

Postby Pfhorrest » Sat Feb 24, 2018 7:38 pm UTC

Kit, I'd like to give you the benefit of the doubt that you're having honest trouble understanding, but it really seems like you're pointedly misunderstanding the conventional sense of "mapping" that Jose is using, taking it instead to mean a more abstract sense something like "building a model of the world". The stuff that cars do with lidar, radar, and cameras is equivalent to what humans do with their eyes, and we don't call that activity "mapping" when a human does it, in the context of driving at least. Cognitive scientists might talk about how vision involves constructing an abstract map of the world, and computer scientists might talk about machine vision (from cameras, radar, lidar, etc) doing the same thing. But that's clearly not the sense involved in this conversation, except on your end.

Jose is obviously talking about cars driving by using a remotely-supplied map of where roads are, and a GPS signal to tell where itself is, and navigating based entirely on that; and he's saying that that's a terrible idea and should not be done, but rather the cars should be using machine vision to gather data about its immediate surroundings in real time and maneuver primarily based on that, with external maps of the larger road structure as a secondary planning tool, like humans do. I think you agree on that practical upshot, and it really seems like you're just insisting that that also counts as "mapping" for the sake of creating disagreement where there is none.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

Mikeski
Posts: 1034
Joined: Sun Jan 13, 2008 7:24 am UTC
Location: Minnesota, USA

Re: 1958: "Self-Driving Issues"

Postby Mikeski » Sat Feb 24, 2018 8:19 pm UTC

Kit. wrote:One flight on a modern airliner is about as dangerous as 10 km by a car.

I assume cars would be similarly safe if they were guaranteed at least a half-kilometer of empty space in every direction*. (Except when they were right by their own garage, where they would only be guaranteed no other vehicles or pedestrians would get in their way, but they would have to avoid stationary objects.)

The safety of commercial airliners (and autopilot) isn't really comparable to the self-driving car problem. Aircraft are safe because they're effectively always "the only car on the road". Wikipedia lists 13 "mid-air collisions" worldwide since 2000. One of those was at an aerobatics show, and one was between two satellites in orbit, which I think should be recategorized as a "mid-no-air collision". :mrgreen:

Give cars those same constraints, and they'll be really safe! I'm just not sure about having only two or three vehicles operating in Minneapolis at any one time, and keeping all the pedestrians indoors while they are...

* - that's for aircraft at different altitudes. Aircraft at the same altitude on different flight lines maintain a 5 nautical mile distance (9.26 km). If two aircraft are on the same flight line, the trailing craft must stay 15 minutes behind.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3576
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1958: "Self-Driving Issues"

Postby Soupspoon » Sat Feb 24, 2018 8:27 pm UTC

Mikeski wrote:
Kit. wrote:One flight on a modern airliner is about as dangerous as 10 km by a car.

I assume cars would be similarly safe if they were guaranteed at least a half-kilometer of empty space in every direction*. (Except when they were right by their own garage, where they would only be guaranteed no other vehicles or pedestrians would get in their way, but they would have to avoid stationary objects.)

Don't forget turtles!

Kit.
Posts: 1079
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Kit. » Sat Feb 24, 2018 8:40 pm UTC

Pfhorrest wrote:Jose is obviously talking about cars driving by using a remotely-supplied map of where roads are, and a GPS signal to tell where itself is, and navigating based entirely on that; and he's saying that that's a terrible idea and should not be done,

That's a terrible and useless idea, as GPS has neither accuracy nor availability for that. It's not "should not be done", it's just not being done at all.

Pfhorrest wrote:but rather the cars should be using machine vision to gather data about its immediate surroundings in real time and maneuver primarily based on that, with external maps of the larger road structure as a secondary planning tool, like humans do. I think you agree on that practical upshot,

There are too points in which I disagree:

1. I'd say that a car without network connection might still operate in some sort of "degraded" mode, but it will need to slow down to achieve the same safety level as with knowledge sharing between cars. But that's a minor one.

2. That "using machine vision to gather data about its immediate surroundings in real time and maneuver primarily based on that" for a car is not given for granted, but needs training. A lot of training with a lot of data. The data is gathered by car sensors, the training happens in separate data centers, the new version of card software, if proven to be safer, then is deployed on the cars. Trying to avoid "surveillance" by requiring cars to only rely on current sensor data won't work because in order to "only rely on sensor data", cars need to survey huge amounts of data to be trained on.

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Sat Feb 24, 2018 8:49 pm UTC

Kit. wrote:What does it have to do with "GPS mapping"?
Read Pfhorrest's reply.

Kit. wrote:Is lane changing "navigation" or "control" in your terms, and why?
Lane changing (itself) is "control", for the same reason that the act of making a right turn without modifying the fender is also "control". Having decided what you want the car to do (change lanes), you are applying input based on existing road conditions to achieve that goal, while keeping everyone's pants clean. Now, the decision to make a right turn in the first place (or to change lanes in the first place) might be based on navigational input (wanting to get to Von's, and Von's is at the upcoming exit, and you need to be in the right lane to accomplish the exit) may be prompted by navigation, but it might also be prompted by vehicle control (a car ahead of you slammed on the brakes, and this is the best way to preserve the paint job). But in any case the actual accomplishment of the lane change is "control", because the it itself doesn't require knowing one's location in the larger sense.

Kit. wrote:
ucim wrote:suppose you [...] visited a car dealership [for a] test drive...
That's already navigation.
The salesman salesperson retail experience manager is doing the navigation. You are merely engaged in vehicle control.
Kit. wrote:Besides, car salesmen prefer to choose safer routes.
Huh? What would be safer than "make a right out of the driveway, go to the light, and then take a left...", and how would you know this?

Anyway, the thrust of what I'm saying is that there is a push for more (and more accurate) GPS mapping to serve as a better database for self driving cars, "because they need it", and that's what I'm calling BS on. Drivers, robotic or otherwise, need to know where nearby things are, not where all things should be. Present mapping (in the road map sense) systems are fine for robotic cars. It's the vehicle control (deciding what to do based on local sensory input) that needs work. And for that, super-detailed GPS mapping is irrelevant.

I also acknowledge that sharing driving experiences between robotic vehicles can make them all better drivers, and is probably a Good Idea. But it can also be used for espionage, in the same way that cookies, scripts, service workers, and apps already do forto us on the web. And this is probably a Bad Idea.

The information is valuable, and self-driving companies are powerful. We can't stop the Bad parts without being aware of them first, agreeing that those parts are Bad second, and figuring out how to prevent robotic cars (and Alexa and Echo and robots in general) from doing these Bad Things third.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

Kit.
Posts: 1079
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Kit. » Sat Feb 24, 2018 9:50 pm UTC

ucim wrote:
Kit. wrote:What does it have to do with "GPS mapping"?
Read Pfhorrest's reply.

I don't see anything that I could recognize as "GPS mapping" there.

Can you give a definition?

ucim wrote:
Kit. wrote:Is lane changing "navigation" or "control" in your terms, and why?
Lane changing (itself) is "control", for the same reason that the act of making a right turn without modifying the fender is also "control".

I'm not talking about "acts", I'm talking about decisions. What are your criteria whether a decision belongs to the "navigation" system or to the "control" system, and what results are produced by applying these criteria to a decision to change lanes?

ucim wrote:
Kit. wrote:
ucim wrote:suppose you [...] visited a car dealership [for a] test drive...
That's already navigation.
The salesman salesperson retail experience manager is doing the navigation. You are merely engaged in vehicle control.

Our notions of what navigation is obviously differ. That's why these criteria above could be useful to continue meaningful discussion.

ucim wrote:
Kit. wrote:Besides, car salesmen prefer to choose safer routes.
Huh? What would be safer than "make a right out of the driveway, go to the light, and then take a left...", and how would you know this?

If the car saleswhatever picked that route, it might already be safer than the alternatives, which means safer than average around. But usually, right turns are safer than left turns.

ucim wrote:Anyway, the thrust of what I'm saying is that there is a push for more (and more accurate) GPS mapping to serve as a better database for self driving cars, "because they need it", and that's what I'm calling BS on. Drivers, robotic or otherwise, need to know where nearby things are, not where all things should be. Present mapping (in the road map sense) systems are fine for robotic cars. It's the vehicle control (deciding what to do based on local sensory input) that needs work. And for that, super-detailed GPS mapping is irrelevant.

Unfortunately, I cannot give a meaningful answer because I don't understand what you mean by "GPS mapping".

But I can try to mention that both humans are cars are working with probabilistic information, sensors don't give absolutely precise data but just improve the knowledge of probabilities compared to the a priori knowledge, and the better the a priori knowledge is, the better the a posteriori knowledge will be.

ucim wrote:I also acknowledge that sharing driving experiences between robotic vehicles can make them all better drivers, and is probably a Good Idea. But it can also be used for espionage, in the same way that cookies, scripts, service workers, and apps already do forto us on the web. And this is probably a Bad Idea.

The information is valuable, and self-driving companies are powerful. We can't stop the Bad parts without being aware of them first, agreeing that those parts are Bad second, and figuring out how to prevent robotic cars (and Alexa and Echo and robots in general) from doing these Bad Things third.

Yeah, it's easier to blame technology for being "a bad thing" than to keep "bad people" from abusing it. The problem is, you cannot force your competitors (China, for example) to do the same, so, by not crippling their technological progress, they may get a competitive advantage over you. Google is already expanding their AI research into China, even though they still have troubles with providing their core services there.

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Sun Feb 25, 2018 12:13 am UTC

Kit. wrote:I don't see anything that I could recognize as "GPS mapping" there.
Can you give a definition?
[...]
Our notions of what navigation is obviously differ. That's why these criteria above could be useful to continue meaningful discussion.
There are two senses in which the word "mapping" is used here. What I'm referring to as "GPS mapping" is the construction of a representation of the external world as it appeared at some point in the past, which was constructed at some point in the past, independently of the use to which is would be presently put. Typically it would be used for navigation. The predominant form nowadays is a set of datapoints georeferenced to GPS coordinates, displayed on a screen which human drivers can refer to, or stored in a data structure which robotic drivers can refer to. An older form involved paper and ink drawings reproduced on a printing press and often accordion-folded to make them easy to store.

For this purpose, navigation is the act of making strategic plans and decisions as to which course to follow in order to arrive at a desired location. This typically will involve the choice of roads to follow, intersections to turn at, and sometimes time-and-distance calculations to help the driver decide when to depart. Navigation is typically pre-planned, but can also be updated on the fly when conditions warrant.

The other sense of "mapping" which is used here (but not by me) is the system by which an entity is aware of its actual surroundings, taking in sensor data, and processing it in order to figure out where things are around it. Sensor data is remembered, and the relative positions of objects in its field are recalculated as needed when the entity moves and has not updated its sensor data. People create mental maps of rooms they are in; they can use these mental maps to navigate for a limited time when the lights go out or they shut their eyes. Bats create a mental map of their surroundings based on echolocation. A "mental map" in this sense is largely another name for a kind of self-awareness.

Kit. wrote:I'm not talking about "acts", I'm talking about decisions. What are your criteria whether a decision belongs to the "navigation" system or to the "control" system, and what results are produced by applying these criteria to a decision to change lanes?
In short, control is a set of acts. Navigation is a set of decisions. Actions of course require decisions (how far to turn the wheel, when to start turning back...) but the primary focus is the action of changing lanes without hitting anything. That is "control". It is a tactical thing.

Navigation is a strategic thing - it is the... er... act of planning a route to a destination. I am not using the word to refer to dodging ("navigating around") an obstacle.

Another distinction is that navigation (as I'm using the term) uses outdated nonlocal information for planning purposes. The map is compiled beforehand, just like a paper map is. An out-of-date map could get you lost, but it shouldn't get you into an accident.

Control (as I'm using the term) uses up-to-date local information for collision avoidance. Sensors provide this information, which includes exactly where the curb is, how wide the lines are, where the nearby vehicles and other obstacles are, and the location (and existence!) of the pothole in front of you. Poor control isn't likely to get you lost, but it could easily get you killed.

There seems to be a widespread feeling that we need "better maps" (of the GPS or paper variety) in order for autonomous driving to be successful, and therefore the (i.e. google) mapping vehicles need to gather more info. I say no, the maps are fine. We need better sensor interpretation. Better "mental maps", as it were.

While I'm comfortable with my neighbor knowing when I ride my bike and what my route is, I am absolutly not comfortable sharing that information with google, amazon, waymo, uber, lyft, Tesla, GM... They have no business knowing my life, and the argument "it will make the robots drive better" is not at all compelling.

===

Kit. wrote:Yeah, it's easier to blame technology for being "a bad thing" than to keep "bad people" from abusing it. The problem is, you cannot force your competitors (China, for example) to do the same, so, by not crippling their technological progress, they may get a competitive advantage over you. Google is already expanding their AI research into China, even though they still have troubles with providing their core services there.
Yes, and this is a bigger problem than the sliver of it I bring up here. And in any case, I'm not blaming technology for being a "bad thing", I am saying that self-driving cars come with technology that is easily used for bad things by powerful entities, and these powerful entities are trying (and succeeding) in convincing us that "these aren't the droids you're looking for". We are, for example, inviting 24-7 surveillance machines into our homes, and telling them all our secrets. Echo, Alexa, Nest, all of them are "ooh, shiny" ways to give control of our data to the companies behind them; it's a land grab that's happening before we realize there's land to grab. Does nobody realize the implications of the disappearance of local sync? But that's another rant for another day. :)

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
Crissa
Posts: 291
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Crissa » Sun Feb 25, 2018 9:13 am UTC

I have driven thousands of miles on roads I did not know the turns on. Of course I grew up before GPS. It's probably not in the thousands of thousands, but easily in the tens of thousands. There were over a hundred miles of paved road in my home town and I drove them all. (I also bicycled a large number of them, but that's neither here nor there, the vast amount of traveling is on roads I've driven before.)

But even I know that a guy telling you directions is a type of map. An annoying one, but a map. Memory is a map.

Google cars memorizing the layout of streets and signs and curbs and traffic densities and trees is a map. Furthermore, all Google cars share the same repository of this memory.

Memory is a map. At least when we're talking about AI.

Mapping is literally memorizing a path or series of landmarks.

Navigation is literally choosing a path, be it the first time I 'navigate' a set of flagstones I've never seen or 'navigate' by using dead reckoning.

This is the goofiest misinterpretation of these words.

-Crissa

PS: Before cars, there weren't roads to most people's houses. Having only trains doesn't mean that we wouldn't have delivery trucks - just that those trucks would be either smaller or shorter range and big industrial vehicles would just move slower. Have you ever tried to make a delivery or pickup on a college campus? You end up driving on walking paths. Moving barricades. Up stairs. It's weird, but it can be done.

sonar1313
Posts: 162
Joined: Tue Mar 05, 2013 5:29 am UTC

Re: 1958: "Self-Driving Issues"

Postby sonar1313 » Mon Feb 26, 2018 4:15 am UTC

Crissa wrote:What gets me is...

...None of these 'what ifs' do humans handle safely. Handwritten signs? Most should stop. Do Humans? An AI would. Bad lane markings. You probably should stop. An AI would. Do humans? Bad direction in a construction zone. You're supposed to stop. An AI would.

So why do we care if an AI decides to stop and call for help when these edge cases happen? That's what should happen.

Sensors fouled at 70 mph? The car stops. What does a human driver do if blinded at 70mph? This happens! Why is the AI case unique?


You stop and call for help when the lane markings are bad? I hope not. Humans are actually much better at handling the unexpected than we give them credit for. If faded or difficult-to-see lane markings caused human drivers to stop and call for help, society would largely break down - and that's not even an exaggeration. Miles and miles and miles of otherwise perfectly useful road would become impassable and there would be hundreds of drivers blocking the road and shoulder and clogging up 911 and Onstar something fierce.

As far as fouling the sensors, I suppose it's possible that autonomous cars will eventually have little wipers or something for each and every one of its sensors, because that's what it would need. Humans can and should slow down or stop in whiteout conditions or rain too heavy to see through, but when that ends, a few flicks of the windshield wipers and off you go. Unless you periodically get out and wipe off all the sonar and lidar eyes, though - like, every half-mile, which still renders transportation so awful as to be useless - conditions that would inconvenience a human driver would cripple an AI one.

User avatar
Wee Red Bird
Posts: 187
Joined: Wed Apr 24, 2013 11:50 am UTC
Location: In a tree

Re: 1958: "Self-Driving Issues"

Postby Wee Red Bird » Mon Feb 26, 2018 8:10 am UTC

Old Bruce wrote:Please only do that if you are a passenger [serious-and-concerned-for-everyone's-safe-commute-face emoticon]

Don't worry, I have a dashcam in one of the cars. Unfortunately not the one I usually take through those road works. Though this weekend, it recorded it's fair share of crashed cars on my travels. None serious for the humans, but the poor cars won't be going anywhere under their own power.

User avatar
Steve the Pocket
Posts: 689
Joined: Mon Apr 23, 2007 4:02 am UTC
Location: Going downtuuu in a Luleelurah!

Re: 1958: "Self-Driving Issues"

Postby Steve the Pocket » Mon Feb 26, 2018 3:13 pm UTC


Upon further reflection, I think I hate that they did this? They spent all this time, money and effort setting up this experiment on just six test subjects, after which they presumably tore everything down again. So they've actually learned nothing useful. And what's worse, they've basically made it impossible for anyone to conduct this experiment ever again because the test subjects are liable to have seen this video and know what's up. Grats, Vsauce, your zeal for Internet views killed any chances of ever getting useful data on this subject.
cephalopod9 wrote:Only on Xkcd can you start a topic involving Hitler and people spend the better part of half a dozen pages arguing about the quality of Operating Systems.

Baige.

Leovan
Posts: 93
Joined: Mon Dec 05, 2016 9:31 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Leovan » Mon Feb 26, 2018 3:14 pm UTC

Radar can penetrate snow and rain and fog, so that's a plus. It's no LIDAR, but avoiding stopped cars in front of them in the fog is something self driving cars should be able to do better than humans.

HERE maps is proposing using the parking assist sonar/radars in cars passing by to identify empty parking spaces, so your car always knows the nearest parking space. So no need to have your car circle the block, it knows where to park. Of course this service is independent of the car being self-driving. I figure a lot of these services being implemented can do both.

So self-driving taxis should be a thing this year... The number of LIDARs sold have also increased dramatically, appropriate for a fleet buildup, not just Waymo. Of course this is Arizona, so no bad weather, and when the weather IS bad, the taxi fleet just stays home and you call a human driven company. They estimate a self-driving car costs about 200k USD, and to have them amortized in 2 years.
https://www.eetimes.com/document.asp?doc_id=1332987
https://arstechnica.com/cars/2018/02/ro ... e-in-2018/

Kit.
Posts: 1079
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Kit. » Mon Feb 26, 2018 7:05 pm UTC

Mikeski wrote:
Kit. wrote:One flight on a modern airliner is about as dangerous as 10 km by a car.

I assume cars would be similarly safe if they were guaranteed at least a half-kilometer of empty space in every direction*.

Can't we just say them that humans are too bad at collision avoidance, and we would like autonomous cars to be better, much better at it?

ucim wrote:There are two senses in which the word "mapping" is used here. What I'm referring to as "GPS mapping" is the construction of a representation of the external world as it appeared at some point in the past, which was constructed at some point in the past, independently of the use to which is would be presently put.

I don't know of those. Normally, maps are constructed for some kind of use, even if this use is quite broad. For example, for use by humans for navigation.

Besides, it has nothing to do with GPS. Probably you wanted to say "static mapping" or something like that. And even if you want to specify a particular coordinate system for those maps, it's not "GPS", it's "WGS 84".

ucim wrote:
Kit. wrote:I'm not talking about "acts", I'm talking about decisions. What are your criteria whether a decision belongs to the "navigation" system or to the "control" system, and what results are produced by applying these criteria to a decision to change lanes?
In short, control is a set of acts. Navigation is a set of decisions. Actions of course require decisions (how far to turn the wheel, when to start turning back...) but the primary focus is the action of changing lanes without hitting anything. That is "control". It is a tactical thing.

Navigation is a strategic thing - it is the... er... act of planning a route to a destination. I am not using the word to refer to dodging ("navigating around") an obstacle.

That's a lot of words for the lack of answer to my question. You are basically saying that there are two ways to look at a decision/action, and whether it is "navigation" or "control" depends on your personal preferences for the way of looking. As you don't specify your personal preferences exactly, I still don't know if lane changing is "navigation" or "control".

Kit. wrote:Another distinction is that navigation (as I'm using the term) uses outdated nonlocal information for planning purposes. The map is compiled beforehand, just like a paper map is. An out-of-date map could get you lost, but it shouldn't get you into an accident.

Control (as I'm using the term) uses up-to-date local information for collision avoidance. Sensors provide this information, which includes exactly where the curb is, how wide the lines are, where the nearby vehicles and other obstacles are, and the location (and existence!) of the pothole in front of you. Poor control isn't likely to get you lost, but it could easily get you killed.

So, when I decide to move to the left lane because I know from the map that I would benefit from making a left turn soon and because I see there is currently an opening in the left lane traffic, it's neither navigation nor control.

What is it then?

Kit. wrote:There seems to be a widespread feeling that we need "better maps" (of the GPS or paper variety) in order for autonomous driving to be successful, and therefore the (i.e. google) mapping vehicles need to gather more info. I say no, the maps are fine. We need better sensor interpretation.

And how do you know that? Are you an expert in sensor fusion?

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Mon Feb 26, 2018 8:22 pm UTC

Kit. wrote:Probably you wanted to say "static mapping" or something like that. And even if you want to specify a particular coordinate system for those maps, it's not "GPS", it's "WGS 84".
This is why autonomous cars will never work.

"Alexacar, take me to Bob's house"
"There are many Bobs in the world. Even considering only the ones that don't require swimming, you need to be more precise"
"Alexacar, take me to Bob Smith's house - the one I do yoga with and called last Monday on the phone."
"You do not do yoga with a house, you do yoga with a person. Also, although you can do yoga in a house, you never have. You've done it at the Kripalu Center, which is a commercial building."
"Alexcar, do you know Bob Smith, the person I do yoga with?"
"Yes, of course. He has a red bicycle, jogs from 6:30 am to 7:30 am, lives alone, works at the JayCee, buys peanut butter but no jelly, doesn't like..."
"Ok Alexcar, do you know where he lives?"
"Yes. I have the entire history of his domiciles at my fingertips. He was born to..."
"Never mind that. Alexcar, do you know where he lives now?"
"Of course I do. It's 45 Albion Lane, New Brun...."
"ALEXACAR - TAKE ME TO HIS HOUSE."
"I'm sorry, I can't do that. He doesn't actually own a house. He's been renting for his entire adult life because his credit rating isn't...."
"Alexacar, never mind. I'll walk."
"Walking is good exercise. I approve. Thank you for using Amazon Driving Services. ADS is always happy to help you out. And we already know which way you came in."

Kit. wrote:As you don't specify your personal preferences exactly, I still don't know if lane changing is "navigation" or "control".
I suppose you'll have to live your life in ignorance.

Kit. wrote:So, when I decide to move to the left lane because I know from the map that I would benefit from making a left turn soon and because I see there is currently an opening in the left lane traffic, it's neither navigation nor control.
When you decide to prioritize moving to the left lane, that's navigation. When you actually turn the wheel, that's control.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3576
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1958: "Self-Driving Issues"

Postby Soupspoon » Mon Feb 26, 2018 9:51 pm UTC

Leovan wrote:Radar can penetrate snow and rain and fog,
...and pedestrians/animals, trees/other vegetation, many rocks, duned sand, snowdrifts, that part-load of PVC piping that accidentally came loose from the flatbed lorry that passed by shortly before..?

rmsgrey
Posts: 3434
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1958: "Self-Driving Issues"

Postby rmsgrey » Mon Feb 26, 2018 10:56 pm UTC

Kit. wrote:
ucim wrote:
Kit. wrote:I'm not talking about "acts", I'm talking about decisions. What are your criteria whether a decision belongs to the "navigation" system or to the "control" system, and what results are produced by applying these criteria to a decision to change lanes?
In short, control is a set of acts. Navigation is a set of decisions. Actions of course require decisions (how far to turn the wheel, when to start turning back...) but the primary focus is the action of changing lanes without hitting anything. That is "control". It is a tactical thing.

Navigation is a strategic thing - it is the... er... act of planning a route to a destination. I am not using the word to refer to dodging ("navigating around") an obstacle.

That's a lot of words for the lack of answer to my question. You are basically saying that there are two ways to look at a decision/action, and whether it is "navigation" or "control" depends on your personal preferences for the way of looking. As you don't specify your personal preferences exactly, I still don't know if lane changing is "navigation" or "control".

Kit. wrote:Another distinction is that navigation (as I'm using the term) uses outdated nonlocal information for planning purposes. The map is compiled beforehand, just like a paper map is. An out-of-date map could get you lost, but it shouldn't get you into an accident.

Control (as I'm using the term) uses up-to-date local information for collision avoidance. Sensors provide this information, which includes exactly where the curb is, how wide the lines are, where the nearby vehicles and other obstacles are, and the location (and existence!) of the pothole in front of you. Poor control isn't likely to get you lost, but it could easily get you killed.

So, when I decide to move to the left lane because I know from the map that I would benefit from making a left turn soon and because I see there is currently an opening in the left lane traffic, it's neither navigation nor control.

What is it then?


The intention to get into the left lane as preparation for a left turn is a strategic decision.
The decision to get into the left lane at a particular moment because there's a suitable opening is a tactical decision.

You don't need to get that complicated to cause a similar apparent dilemma between whether something is navigation or control - just taking a corner does it. The decision that you want to take that corner is a strategic/navigational one, while the decisions on when to turn the wheel and how far in order to get round the corner are tactical/control ones. Pretty much every decision taken when driving has both navigational and control (or both strategic and tactical) aspects - on the navigational level, you have a goal of reaching your destination, and subgoals that set out your route to various levels of precision; on the control level, you have the actions to take in order to achieve your next subgoal, while also not hitting anything. Wanting to get to the end of the street is a navigational goal; not swerving onto the sidewalk is a matter of control.

User avatar
riker1384
Posts: 2
Joined: Tue Nov 13, 2012 5:18 am UTC

Re: 1958: "Self-Driving Issues"

Postby riker1384 » Fri Mar 02, 2018 7:30 am UTC

Personally I think putting yourself in a self-driving car is insane. You're putting yourself in a death-box that can kill you at the whim of its software. Would you be comfortable sitting in front of a loaded gun controlled by an application running on your Internet-connected desktop computer? I wouldn't. Airplanes do have fly-by-wire software, but it's very highly vetted, isolated from the outside, and as mentioned above flying is much simpler than driving on the ground.

Aside from the machine-vision issues discussed above, the other thing that really worries me is hacking. Imagine that somebody puts some obfuscated malicious code into a software update. It could be an external hacker, or it could be an employee of the car company who is disgruntled or who is acting as an agent of a foreign power. (Let's say we get in a war with China or Iran, and they have a Tesla programmer working for their intelligence service.) At some particular time and date, every car on the road accelerates and crashes into the nearest car or obstacle at maximum speed. You could potentially kill millions of people, more than you could with one nuclear bomb, and wreck the economy because everyone would be scared to drive to work the next day.

How can they guarantee this won't happen? The probability might be low but the consequences are catastrophic. If Stuxnet managed to hack into Iran's nuclear fuel enrichment plant, I don't believe a car company is going to be more secure than that. And as I mentioned, there's the threat of internal sabotage by the programmers. People do crazy and commit massacres from time to time, and spies and saboteurs do exist. The government and military are penetrated by spies all the time, and a private car company wouldn't even be held to their standards.

There's also the threat of targeted attacks, murder and assassination of individuals. Some hacker, or someone in the car company or the government gets code written into your car to crash and kill you. (Or if they really want to be mean, they overheat the motors and batteries so you burn to death, and lock you in because apparently Tesla thinks electronic door latches are a good idea.) There's not a damn thing you can do about it.

I feel the same way about any car that has drive-by-wire capability without direct control over steering and brakes, even if it doesn't have full self-driving yet. There should always be a manual override of the steering, brakes and power. Of course, in a fully self-driving car even that won't help you if you're taking nap when it decides to kill you.

User avatar
HES
Posts: 4872
Joined: Fri May 10, 2013 7:13 pm UTC
Location: England

Re: 1958: "Self-Driving Issues"

Postby HES » Fri Mar 02, 2018 10:24 am UTC

riker1384 wrote:Personally I think putting yourself in a self-driving car is insane. You're putting yourself in a death-box that can kill you at the whim of its software.

Do you ever get in a taxi?
He/Him/His Image

User avatar
GlassHouses
Posts: 129
Joined: Thu Nov 24, 2016 12:41 pm UTC

Re: 1958: "Self-Driving Issues"

Postby GlassHouses » Fri Mar 02, 2018 12:45 pm UTC

HES wrote:
riker1384 wrote:Personally I think putting yourself in a self-driving car is insane. You're putting yourself in a death-box that can kill you at the whim of its software.

Do you ever get in a taxi?

That's different. The taxi driver is trying to stay alive, just like you are. Unless you find yourself with a suicide cabbie, but I think those are quite rare. (Also, taxi drivers are at far, far greater risk of being murdered by their passengers than vice versa.)

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Fri Mar 02, 2018 4:28 pm UTC

HES wrote:Do you ever get in a taxi?
Not one whose driver is part of the hive. The networked aspect of AI cars is a significant problem. Driving should be local, not global. Just like voting.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

speising
Posts: 2272
Joined: Mon Sep 03, 2012 4:54 pm UTC
Location: wien

Re: 1958: "Self-Driving Issues"

Postby speising » Fri Mar 02, 2018 5:53 pm UTC

considering win10's update policy, and also the latest rant in the rant thread viewtopic.php?f=2&t=22551&p=4321278#p4321278 , i'd be somewhat afraid of receiving a compulsory firmware update while doing 130 on the highway, complete with non postponable reboot cycle.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26453
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1958: "Self-Driving Issues"

Postby gmalivuk » Sun Mar 04, 2018 2:46 pm UTC

riker1384 wrote:Personally I think putting yourself in a self-driving car is insane. You're putting yourself in a death-box that can kill you at the whim of its software. Would you be comfortable sitting in front of a loaded gun controlled by an application running on your Internet-connected desktop computer? I wouldn't. Airplanes do have fly-by-wire software, but it's very highly vetted, isolated from the outside, and as mentioned above flying is much simpler than driving on the ground.

Aside from the machine-vision issues discussed above, the other thing that really worries me is hacking. Imagine that somebody puts some obfuscated malicious code into a software update. It could be an external hacker, or it could be an employee of the car company who is disgruntled or who is acting as an agent of a foreign power. (Let's say we get in a war with China or Iran, and they have a Tesla programmer working for their intelligence service.) At some particular time and date, every car on the road accelerates and crashes into the nearest car or obstacle at maximum speed. You could potentially kill millions of people, more than you could with one nuclear bomb, and wreck the economy because everyone would be scared to drive to work the next day.

How can they guarantee this won't happen? The probability might be low but the consequences are catastrophic. If Stuxnet managed to hack into Iran's nuclear fuel enrichment plant, I don't believe a car company is going to be more secure than that. And as I mentioned, there's the threat of internal sabotage by the programmers. People do crazy and commit massacres from time to time, and spies and saboteurs do exist. The government and military are penetrated by spies all the time, and a private car company wouldn't even be held to their standards.

There's also the threat of targeted attacks, murder and assassination of individuals. Some hacker, or someone in the car company or the government gets code written into your car to crash and kill you. (Or if they really want to be mean, they overheat the motors and batteries so you burn to death, and lock you in because apparently Tesla thinks electronic door latches are a good idea.) There's not a damn thing you can do about it.

I feel the same way about any car that has drive-by-wire capability without direct control over steering and brakes, even if it doesn't have full self-driving yet. There should always be a manual override of the steering, brakes and power. Of course, in a fully self-driving car even that won't help you if you're taking nap when it decides to kill you.
Your car can probably already be hacked, btw.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
Bloopy
Posts: 213
Joined: Wed May 04, 2011 9:16 am UTC
Location: New Zealand

Re: 1958: "Self-Driving Issues"

Postby Bloopy » Wed Mar 07, 2018 12:33 pm UTC

gmalivuk wrote:
DataGenetics wrote:TL;DR - If you know a car is going to stop when you step into the street, why not simply step into the street whenever you like?
You can already be reasonably sure (at least here) that most human drivers will try to stop for pedestrians, whether or not the pedestrians have legal right-of-way. Physics is still an issue, though.

If a pedestrian looks like they might be about to cross the road, human drivers will be more cautious around them. The challenge is to suddenly change direction and leap out onto the road unexpectedly, catching a driver who was looking for an address or poking a dashboard button off guard. The game just won't be the same once humans give up the wheel.

User avatar
bantler
Posts: 238
Joined: Thu Apr 08, 2010 11:23 pm UTC

Re: 1958: "Self-Driving Issues"

Postby bantler » Wed Mar 07, 2018 10:20 pm UTC

Bloopy wrote:If a pedestrian looks like they might be about to cross the road, human drivers will be more cautious around them.


I find if cars look like they are not cautiously slowing down to stop, pedestrians stay on the sidewalk.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3576
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1958: "Self-Driving Issues"

Postby Soupspoon » Wed Mar 07, 2018 11:07 pm UTC

bantler wrote:
Bloopy wrote:If a pedestrian looks like they might be about to cross the road, human drivers will be more cautious around them.


I find if cars look like they are not cautiously slowing down to stop, pedestrians stay on the sidewalk.

Conversely, I've had cars that I've carefully assessed the speed of, such that I will cross their side of the road well after they've passed by1, suddenly notice me on not-a-collision-course and then put their anchors on, to let me cross in front. Needlessly. And occasionally with arm-waving to indicate that they thought that I wasn't paying attention, when I could honestly say that they must have been distracted2 to have so reacted.


1 Although I don't mind almost scraping my nose on the rear end of a bus as it passes by, say, to give me more time before the next vehicle (itself many vehicle lengths behind, but still moving steadily with no chance of surprises) needs me to vacate their forward comfort-zone.

2 As a driver, myself, I know there's some statutory time necessary to glance at the mirrors and elsewhere (if you're doing it right), but that doesn't explain it.

User avatar
Pfhorrest
Posts: 4888
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1958: "Self-Driving Issues"

Postby Pfhorrest » Wed Mar 07, 2018 11:33 pm UTC

I've had that experience too, and also the opposite experience as a driver. Someone way far ahead of me will enter a crosswalk. I will adjust my speed down to make sure that I won't be near the crosswalk until they're through it. But then they slow down, so I slow down more. And then they slow down more, and I slow down more. And eventually they're stopped in the middle of the crosswalk staring at me like to make sure I'm not going to run them over at 1mph. Dude, just fucking cross the street already! You'd have been perfectly safe just crossing like you were, and now we're both stopped in the middle of the road staring at each other. Go!
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
Steve the Pocket
Posts: 689
Joined: Mon Apr 23, 2007 4:02 am UTC
Location: Going downtuuu in a Luleelurah!

Re: 1958: "Self-Driving Issues"

Postby Steve the Pocket » Fri Mar 09, 2018 2:59 am UTC

Clearly the optimal solution is to implement computer-controlled pedestrians as well.
cephalopod9 wrote:Only on Xkcd can you start a topic involving Hitler and people spend the better part of half a dozen pages arguing about the quality of Operating Systems.

Baige.

Mikeski
Posts: 1034
Joined: Sun Jan 13, 2008 7:24 am UTC
Location: Minnesota, USA

Re: 1958: "Self-Driving Issues"

Postby Mikeski » Fri Mar 09, 2018 6:26 am UTC

Steve the Pocket wrote:Clearly the optimal solution is to implement computer-controlled pedestrians as well.

It's been done.

I've seen the pedestrians with the forward-mounted touchscreens. They seem even more oblivious to traffic than the non-computerized ones, though.

User avatar
Archgeek
Posts: 186
Joined: Wed May 02, 2007 6:00 am UTC
Location: Central US
Contact:

Re: 1958: "Self-Driving Issues"

Postby Archgeek » Mon Apr 23, 2018 7:27 pm UTC

ucim wrote:We are, for example, inviting 24-7 surveillance machines into our homes, and telling them all our secrets. Echo, Alexa, Nest, all of them are "ooh, shiny" ways to give control of our data to the companies behind them; it's a land grab that's happening before we realize there's land to grab. Does nobody realize the implications of the disappearance of local sync? But that's another rant for another day. :)

Jose


I am right there with you vis-a-vis the "smart speaker" infiltration. I'm a computer engineer who knows the real workings of AI to the point that I'm not afraid of it, but those toys give me the dang creeps. I must know, however, as I've not encountered the phrase within recollection, what precisely is local sync? I sense it's either something I already know but not by that name, or something I never used due to actively avoiding its platform (e.g. alien space rectangles cell phones until about two years ago).
"That big tube down the side was officially called a "systems tunnel", which is aerospace contractor speak for "big tube down the side."

rmsgrey
Posts: 3434
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1958: "Self-Driving Issues"

Postby rmsgrey » Tue Apr 24, 2018 1:18 am UTC

Archgeek wrote:what precisely is local sync?


My assumption is that it's like cloud sync, but with a local store rather than sending data across the internet.

So files or directories with local sync enabled will be automatically backed up to a designated mirror either attached to the same machine, or on the same LAN.

I could be wrong, but it's consistent with what comes up from a cursory Google.

User avatar
ucim
Posts: 6488
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1958: "Self-Driving Issues"

Postby ucim » Mon Apr 30, 2018 3:15 am UTC

Archgeek wrote:what precisely is local sync?
Perhaps I should have said "direct sync". It means synchronizing with a local machine, locally, without going through the cloud. The Palm does that; you connect the Palm physically (or optically, or bluetoothly) with a computer and copy the information across that connection. Nobody on the internet ever needs to know (except for any leakage from the optics or the blue teeth).

It's also useful to have your data in the cloud; your house could burn down and you'd still be able to remember your cousin's birthday because the computer that has your data is off-site. But that's not the only way to have a copy off-site.

The thing that irks me though, is that devices nowadays do not permit local sync. Sure, you can sync to the computer that is right next to you, but you have to use an app, and go through the internet, where the app authors can collect the info for their own purposes and sell their conclusions to anybody who pays them. You can't just sync directly to your local computer.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

Leovan
Posts: 93
Joined: Mon Dec 05, 2016 9:31 pm UTC

Re: 1958: "Self-Driving Issues"

Postby Leovan » Mon Apr 30, 2018 5:26 am UTC

ucim wrote:The thing that irks me though, is that devices nowadays do not permit local sync. Sure, you can sync to the computer that is right next to you, but you have to use an app, and go through the internet, where the app authors can collect the info for their own purposes and sell their conclusions to anybody who pays them. You can't just sync directly to your local computer.

Jose


It's even worse than that. My Fitbit won't even sync with my phone unless my phone has internet.

User avatar
GlassHouses
Posts: 129
Joined: Thu Nov 24, 2016 12:41 pm UTC

Re: 1958: "Self-Driving Issues"

Postby GlassHouses » Mon Apr 30, 2018 10:59 am UTC

ucim wrote:The thing that irks me though, is that devices nowadays do not permit local sync. Sure, you can sync to the computer that is right next to you, but you have to use an app, and go through the internet, where the app authors can collect the info for their own purposes and sell their conclusions to anybody who pays them. You can't just sync directly to your local computer.

iOS devices can still sync locally, no Internet required.


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: mscha and 32 guests