- Google's self-driving cars can tour you around the streets of Mountain View, California.
I know this. I rode in one this week. I saw the car's human operator take his hands from the wheel and the computer assume control. "Autodriving," said a woman's voice, and just like that, the car was operating autonomously, changing lanes, obeying traffic lights, monitoring cyclists and pedestrians, making lefts. Even the way the car accelerated out of turns felt right.
It works so well that it is, as The New York Times' John Markoff put it, "boring." The implications, however, are breathtaking.
Shit, I've been following the whole self-driving car buzz from the start and I didn't know it relied so heavily on pre-registered data. Disappointing, really. It's like finding out your favourite band actually playbacks at concerts. How typical of Google, really: can't solve a problem? Throw big data at it until it works. I was under the assumption that it scanned, processed and calculated everything in real-time - which is what you ultimately need. You don't want to be dependant on an external source. Imagine if, due to some technical problem, your car loses access to that data, while you're traveling at 120 km/h. What happens then? Besides, their data is outdated the second it is created. I read earlier that when roadworks are happening, the car can't cope with it. Now I know why - it's not that it is a much more complex situation, but because the car isn't capable of doing so.
Doing everything in real-time is difficult with current sensor technology and processing algorithms, and leaves a lot of room for error. The way that they're doing it currently is fairly clever for minimizing the amount of calculation/reaction time. I'm not so sure it absolutely needs to be real-time, so long as the system works. That said, I'm no stranger to the 3D CAD universe, and I can tell you that housing all of the spatial data, regardless of format, is a LOT of disk space. A one inch resolution for 2,000 miles of mapped roadway is going to cost you. The best solution might be to operate in "blocks" of spatial data, similar to Minecraft. If the car fails to load the next block, it alerts the occupant, who will be forced to take the reins, or else the car will pull itself off of the road and park near the horizon of its current block until a connection with the server is restored. Naturally, for your native city, you'd just have the whole thing stored locally. For road trips, input a destination and route to download into the car before you left. You might only need the blocks solution if you're driving aimlessly, or didn't have time to input destination/route. You're always going to require a GPS connection or the next-generation equivalent sometime anyway. Sorry, just playin' devil's advocate, like I do. :)
No, that's exactly my point: a self-driving car should be a human replacement, not your GPS that happens to also control the wheel. GPS'es are loaded with data, too, whether it is pre-loaded or daily updated. The way this system seems to work is that it requires more detailed geographical data, which becomes ever more feasible in the near future. It's just a box, like a GPS, that uses a database of locational data to know where it is and make assumptions based on that data. Only now the assumptions aren't 'where should I go' but 'where do I steer this thing'. I understand that the way they do it now is a good stepping stone, but Google being Google, I am afraid that when as soon as they can they will mass-produce these cars and just make money off that. It's a reasonable business case. It is not, however, a real human-AI-like self-driving car. Which is exactly the problem. Normal humans don't depend on a colossal GIS/CAD database to find their way. A normal driver should be able to find its way reading signs and having a general idea of the route to take. I don't see why a self-driving car shouldn't do the same. If self-driving cars can operate without needing preloaded data, you don't need all that disk space at all: just lots of realtime calculations. All the car needs to know for more than a couple of seconds is where to go and the implied road speed. The rest should be handled by the car's AI. That's how I envision self-driving cars. Of course it is difficult, and I'm not saying it should be available now, but I don't see Google going down that path if this trick will put them far ahead of competitors and make them tons of money.You're always going to require a GPS connection or the next-generation equivalent sometime anyway.
and I can tell you that housing all of the spatial data, regardless of format, is a LOT of disk space.
I think this is something to strive for too, but our AI just isn't there yet. Yes, this is essentially a method of compensation... for now. So let's say that we do eventually develop a human-like perception and AI (or better) for this application. What incentive would you have to remove the CAD library functionality? You can argue that it's a crutch right now, but in the future it will be touted as a feature. Indeed, instead of having teams of people with surveying equipment romping around, the sensory data from the cars is inputted into the CAD database. Example: A car changes lanes because 700 ft. ahead (too far for sensor perception) the database knows there's some potholes in the previous lane, as perceived by the car that traveled through the area earlier, right after a heavy downpour. You're trying to implement human perception and processing in a computer (a good idea), but strip it of all of the benefits computers have over us. The two can work together for a system far superior to human-like perception and processing alone. We don't have a perfect track record, after all. And the cost of disk space? Well, we all know what direction that's headed. And... we kinda do depend on a database to find our way, it's just not 3D CAD. And of course, Google already owns the most extensive database (Google Maps). Whatever it takes to make auto-automobiles (heh) a reality as soon as possible... well, I'm for it, as long as you can maintain a failure rate lower than human error. I can understand your disappointment in the impurity/immaturity of the tech though. Appreciating the discussion. :)a self-driving car should be a human replacement
A normal driver should be able to find its way reading signs and having a general idea of the route to take. I don't see why a self-driving car shouldn't do the same. If self-driving cars can operate without needing preloaded data, you don't need all that disk space at all: just lots of realtime calculations
So am I! I'm actually sorta considering writing my thesis about this topic. They aren't mutually exclusive. The important distinction here is what the system is based on: geographical, stored data or sensory input. Google's car is based on stored data, but uses laser input to fill in the blatant gaps in its database, like other cars, people and traffic lights. What I'm mainly arguing is that I think a system based on sensory input, with additional database input, is what Google should be making, but it doesn't appear that they will do so (at least not based on what I've read and what I know about the company. This can change of course). Stored data needn't be abandoned, I just don't want my self-driving car to rely on Google's database. 'Real' self-driving cars (cough notruescotsman) are cars that can drive anywhere, on their own, just like a human being. The benefit here is that the AI can be massively improved by stored data. But I'm thinking more of Waze-like live-reporting of the road's errors, which your self-driving car can then adapt to.Appreciating the discussion. :)
You can argue that it's a crutch right now, but in the future it will be touted as a feature.
Niceeee, I've seen a few maps that you've made... excellent work, of course. And yeah, this would be a pretty fun thesis, are your other choices this cool? Well, then I think that we are in complete agreement. :) I know I've been led astray by Google Maps, and the thought of giving control of death on wheels to a similar database... Google claims impressive statistics right now, but we'll see what happens when this is slowly implemented. Most people won't want to be guinea piggin'. As you probably know, there is a large incentive for other companies (and maybe universities, I'm not sure) to develop the AI, sensory imaging, processing, code that Google then buy$. I dunno about the timescale there, but it can't come soon enough. I'm calling it now though, there will be some not-too-distant movie where someone "HAX A CAR" and commits murder via modifying database files and overriding sensory input or processing. Edit: You taught me the "no true scottman" fallacy today! Cool.I'm actually sorta considering writing my thesis about this topic.
What I'm mainly arguing is that I think a system based on sensory input, with additional database input, is what Google should be making...
It really depends if I can get a good teacher to support me on it. Usually I have to pick from a couple of themes and base my subject on that theme - this year had Sustainable Mobility as one, for example. That might be too complicated for the Michael Bay-fans! We'll probably see a Transformers rehash where the self-driving cars are the villains because they trap humans or whatever. Niceeee, I've seen a few maps that you've made... excellent work, of course. And yeah, this would be a pretty fun thesis, are your other choices this cool?
I'm calling it now though, there will be some not-too-distant movie where someone "HAX A CAR" and commits murder via modifying database files and overriding sensory input or processing.
Edit: You taught me the "no true scottman" fallacy today! Cool.
veen, am_Unition - I enjoyed your conversation greatly. It's interesting the perspective you both bring to it; I gave up on The Age of Context halfway today so I come at it from the perspective of someone who has had "the Internet of Things" shoved down their throat for a day and a half. Veen - you need to do your thesis on this. You could even get a book out of it. There's some serious implications in the key statement here: This is huge and kind of chilling: it implies an "Apple Maps"-grade controversy involving heavy equipment. Consider: 1) An autonomous car using Google's system is at least partially dependent on the quality of a proprietary, licensed dataset. 2) The performance of an extant autonomous vehicle is subject to change based on the licensing in place between the dataset provider and the vehicle manufacturer. 3) The accessibility of a region is dependent on the priority assigned to it by a proprietary, financially-motivated vendor. Imagine Ford comes out with an autonomous vehicle using Google's dataset. Maybe two years later Germany gets into a dust-up with Google over data privacy. Google could retaliate against Germany by downgrading or removing their German dataset - unless Ford has an agreement with Google, in which case Germany could retaliate against Ford. The legal snarl evolves pretty quickly regardless of the rules on the ground. Imagine a private community doesn't want Google Streetview in their development for privacy concerns. Google doesn't map their neighborhood for use with autonomous vehicles. A blind man in another part of town sues the entire development for millions of dollars for violating the Americans with Disabilities Act - after all, the community is preventing him from driving there as autonomous vehicles are an approved assistive device (theoretically). Does privacy trump access? Imagine Audi strikes an deal with Facebook to exclusively use Facebook apps in all its cars. This includes updating all previous Audis from Google Autodrive to Facebook Freedom™. Google argues in court that all Audis from 2015-2022 were qualified as autonomous as a complete system using the Google dataset and that Facebook has not qualified existing Audis with their dataset, thereby requiring a new round of NTSB testing. Facebook countersues but Google gets a temporary injunction - in 9387 of 18433 municipalities. Each one of those municipalities faces different pressures from irate Audi owners that are suddenly forced to pay attention during their entire hour-and-a-half commutes - time they had not considered when they purchased their McMansions 30 miles from their work. How many of them are going to sue? Are they going to sue Audi? Facebook? Google? We're looking at a potential future in which basic access to transportation might very well be contingent on the proprietary products of large corporations. It's much bigger than GPS or no GPS, I think - the fortunes of a community suddenly become dependent on how well they've integrated with their proprietary overlords (and don't for a minute think Google, Apple, Facebook, Amazon, anybody wouldn't charge Podunk, MS through the nose for "premium" autonomous data). And, just as with maps, Google's got a head start to the top of the mountain. Regulations will, most likely, be years behind - how far up the legal ladder will the lawsuit have to go before guilt is assigned in the case of Jon Smith, who died in a tragic fireball because Flatbush NJ allegedly moved a traffic signal without notifying Google and Apple? In the US, at least, we have a Supreme Court that thinks HBO is broadcast over the air. Expecting a nuanced understanding of machine vision, GIS, LIDAR limitations in fog and the value of a pothole map is a pathway to disappointment. I can't think of any other public/private system in which the interplay between a large, financially-motivated monopoly and a basic human right is so primed for strain. The potential implications are, to my mind at least, staggering.The key to Google's success has been that these cars aren't forced to process an entire scene from scratch. Instead, their teams travel and map each road that the car will travel.
Thanks! Was it the book or the idea that made you give up reading it (and if it's the latter, I'd be interested to hear your thoughts on it).
From a legal standpoint, it will be really interesting to see how the responsibilities for this technology will be determined. Is it the software that is responsible for the car's behaviour, or is the driver still expected to keep an eye out and thus still responsible for what the car may do? Will different states and countries respond differently to this? It almost reminds me of scifi movies, where we need to make laws to govern AI's responsibilities. But I'm not versed enough in law to research something as complicated as this will be. What I'm more interested in (and more capable of actually saying something meaningful) are the grave implications such a technology can have on our car-dependant societies. Especially in America, the most car-dependant nation in the world. How is [insert anything car-related] going to change because of this? Could this lead to a decentralised public transport system, one that abandons trains and subways in favour of train-like self-driving cars? Assuming this will be expensive at launch, what will the socio-economical consequences be? Will this fuck our sustainability goals even more, or is this gonna boost electrical cars even more? For more than a hundred years, our average commuting time has stayed roughly the same: will this continue, or will people work more in their car, only appearing in the office an hour per day for a meeting? Maybe minivans become mini-workplaces. Will cabs still exist? How will this affect logistics? When will we trust these cars enough to hand over our lives to lines of code? What is love, baby don't hurt me, don't hurt me, no more? I could go on like this for hours. I'll definitely try to get my thesis on this subject, which I'll have to arrange in eight months or so. I'm not entirely free in my decision. Maybe I should go for a PhD so I can be the first expert in this new area. You never know. Not just disappointment: deaths will happen because of bad nation-wide legislations handling tech like this.veen, am_Unition - I enjoyed your conversation greatly. It's interesting the perspective you both bring to it; I gave up on The Age of Context halfway today so I come at it from the perspective of someone who has had "the Internet of Things" shoved down their throat for a day and a half.
I can't think of any other public/private system in which the interplay between a large, financially-motivated monopoly and a basic human right is so primed for strain.
In the US, at least, we have a Supreme Court that thinks HBO is broadcast over the air. Expecting a nuanced understanding of machine vision, GIS, LIDAR limitations in fog and the value of a pothole map is a pathway to disappointment.
It was the breathless, uninformed cheerleading for the death of privacy. It was the utter certainty that Google Glass was as disruptive a technology as personal computers and if the authors couldn't explain why, it wasn't a failing of their assertion it was just something you'd have to take their word for. It was the slavish adulation for tech for tech's sake, rather than for its ability to solve problems. It starts with a foreword in which an author derides a hotel for not having theater tickets, dinner reservations and a cup of cocoa waiting for him even though he tweeted the hotel's name, his desire for theater tickets, dinner reservations and a cup of cocoa from his plane. It's just a tedious book by tedious people who turned an important subject into utter tedium. Worse, rather than give a broad overview of the subject with positives and negatives, the authors served as moutpieces for anyone who would sit down for an interview with them. It's a sloppy and pointless book. Exactly my point, and exactly what I was trying to convey. Apologies if that wasn't clear. My point about the legal issues is that they're going to be a cast-iron bitch to settle in a fair and beneficial way, which means there are going to be a million different ways the cookie can crumble. I think the world would benefit from an exploration of the implications of a transportation system where key components are locked behind a paywall on a server remote from the vehicle or owner. That's something new under the sun. I don't think you need to be a legal scholar to dig into it - I think you need to be able to look at the impact of a couple different outcomes from a couple different controversial points and then draw some historical parallels in order to get attention. Consider how the destruction of railroads affected the economy of small towns in the United States, or the rise of the Interstate freeway system. The rest of your hypotheticals are totally en pointe. Someone with a Ph. D in, say, "Transportation Information Systems" will be able to write their own paychecks in the next 20 years. Bank on it.Thanks! Was it the book or the idea that made you give up reading it
What I'm more interested in (and more capable of actually saying something meaningful) are the grave implications such a technology can have on our car-dependant societies.
Maybe I should go for a PhD so I can be the first expert in this new area. You never know.
Sounds to me like someone liked his Google Now a bit too much and went on an extrapolating spree. On the one hand, I think predictive knowledge is a really cool feature, and I can envision a future where information is handed to you as soon as you want to know it. I like Google Now, and wish it would work better; it was of great use to me when I flew to Hong Kong, as it has a card with all your flight information. But on the other hand, it's just that: a cool feature. I don't think it is a good foundation to build information technologies on. It is, however, a nice interactive display of information. I think that that book of yours didn't get that right, especially with that hotel example. It was trying to imagine an IT world based around predictive knowledge, instead of seeing the tech for what it's worth: a nice representation of knowledge, a layer on top of current systems. Good way to phrase it. It's exactly why I was disappointed to find out that Google didn't use real self-driving. I don't want my car to be dependant on an external, locked system. There was a great article a while back here, about who owns place. With Google's head start in the mapping business (and to be fair, their maps are the best both cartographically as well as aesthetically), they have a very dominant place in the mapping business. Currently, there are good alternatives, but imagine a similar dominant force in a market with a much higher bar for entry: the radar equipped AI batallions of tech that are Google cars. I'm writing this from a train, sitting only meters from a very capable operator who could be replaced by AI far easier than any car envirnoment. Yet I'm happier that it is just a person. Sidenote: I can't wait to see what Germany thinks of this. They've been the only country to question Google's ethics with streetview. Might be a smart move to do more with GIS. It is already quite up my alley. I wish ArcGIS was up Adobe's level of refinement, though. The thing with this subject is that there is a fairly large chance that a proper Google prototype will never make it to the production stage, simply because states or countries might not allow it. It took crazy ol' California to try it out, after all. Besides, it might face similar perception problems that planes have now: as soon as one crashes, it makes the news. A bad start can halt the development of this tech for decades, easily. If I may extrapolate, that is.It was the breathless, uninformed cheerleading for the death of privacy.
I think the world would benefit from an exploration of the implications of a transportation system where key components are locked behind a paywall on a server remote from the vehicle or owner. That's something new under the sun.
Bank on it.
I think it's worth looking into. For that matter, it might be worth reading the book. As I said, they only reported on people who talked to them, and they were rosy as fuck; that said, they talked to Ford, which apparently plans on consolidating down to economy, luxury and mid-size cars that are going to be largely differentiated by software. My problem with "the Internet of Things" is that it used to be called "telemetry." Telemetry is valuable - it gives you instrumentation for things you need to know. 'The Internet of Things' is predicated on putting sensors on shit and then figuring out how you're going to use them. It's the wrong approach, in my opinion; not everything that sucks power needs an IP address. Somethings definitely benefit though. GM wants to call it "the Industrial Internet" and observes that if telemetry improves the efficiency of their aircraft engines by 1%, it'll save airlines three billion dollars. That's a very different discussion than the "smart fridges" that are generally talked about by breathless tech evangelists after every CES.
Too busy for a long comment, but in a nutshell... Yes. Thank you for framing this in a larger perspective, you nailed it. It's going to be a legal circus, and I'm stocking up on popcorn.
Imagine Ford comes out with an autonomous vehicle using Google's dataset. Maybe two years later Germany gets into a dust-up with Google over data privacy. Google could retaliate against Germany by downgrading or removing their German dataset - unless Ford has an agreement with Google, in which case Germany could retaliate against Ford. The legal snarl evolves pretty quickly regardless of the rules on the ground.
Whatever happened to good ole anti-trust laws? -Seems that would fall squarely in the "abuse of monopoly power" category... right?
Absolutely. Those tend to resolve quickly and without drama. If you want a dizzying and boring read, follow A&P and anti-trust. Long story short, they were Walmart before Walmart and were squarely in the sights of anti-trust litigation until Pearl Harbor, at which point the FDR administration decided they didn't want to mess with groceries during war.
Yeah, but Walmart really can claim that it has direct and viable competitors, right? Who could google point to? Not a lot of mom and pops out there with a small corner store specializing in driverless-car-mapping.
Totally agree. My whole point is that the implications of Google's method are pretty much the definition of "disruptive." There's no real way to predict how this will play out. The conventional view has been "computers are getting smart enough that we won't need to drive anymore." This article (thanks) illustrates that it's more accurate to say "Google wants to be your chauffeur and are willing to turn the entire world into a toll road to do it."
What's interesting is that unlike Walmart (not a lot of love from the high brow) most people will be applauding Googles acquisition of what will become a ubiquitously used product and will not think twice about the implications. Walmart = Evil! Google = innovative :) It will be interesting to see where this leads. I for one cannot wait until I can turn the reigns over to a reliable program to drive for me. It would give me back hours of work time. Side note: Have a spare hour tonight and I plan to send you some files!!