You can already call into question any aspect of anything that's studied at any research institution. There's no need to acquire special justification. Questioning past results is an important part of science. Otherwise we'd have the Sun revolving around the Earth and supposed cold fusion.
That said, one should really have a logical argument based on all the relevant available facts when calling such things into question. Otherwise it's all just a waste of time for everyone involved...particularly the scientists involved, who could be making new discoveries in that time.
In this case, the best one can say is that the idea that 50 million refugees would materialize so soon was wrong, and that Dr. Bogardi of the UNU made a serious mistake in his prediction, which was possibly driven by the desire for creating scary headlines. That last bit about the headlines and anything further is speculation and politics, which is outside the realm of science.
The problem with expanding human rights is that in order for human rights to be meaningful, it must be possible to not trample upon them in any reasonable situation.
If you say that access to food is a human right, then what happens when a supervolcano erupting causes a massive food shortage due to long term lack of light? If we simply don't have enough food, how can we possibly fulfill everyone's right to eat? It comes down to providing everyone insufficient food to survive the disaster (leading to everyone dying) or keeping some people alive by abandoning others early on. Either way we're trampling this right, but the right thing to do is pro-actively trample on it right away so some survive. Would we be monsters for letting people die of starvation when we still had food?
How can we seriously consider something that may reasonably be impossible to fulfill a "right"?
Compare this to the right to free speech. It may be convenient or even beneficial to suppress this right in certain situations, but there's not a situation where it must be suppressed out of physical necessity. Even someone shouting "FIRE" in a crowded theater can be allowed to do so and even allowed to walk free afterwards...
Perhaps what is needed is a new word, or another old word, to describe the sort of "extended rights" that rely on a functioning economic system and the labor of others to provide. Entitlement seems to be popular, but it may fall into the same trap as "right" does...are we all entitled to food in the supervolcano scenario?
Maybe we should turn this on its head and put an emphasis on those providing the benefit to those in need of it? Instead of it being a right for the poor to eat or have access to the Internet, maybe we should see it as an obligation of those with the power to provide these things to do so...a sort of modern noblesse oblige.
While you are technically correct, looking at the data you linked to there are only 8 years over that 65 year time period where it differed from a 17%-20% average: 1949-1951, 1959, 2003-2004, and 2009-2010. If you increase the range to 16%-20% then it falls to a mere 4 years, all of which were in the 14% range: 1949-1950 and 2009-2010. Also, these years tend to be clustered together, which probably indicates they were caused by short term events.
It seems to me that these anomalous years don't shoot down the basic idea that Federal revenue has stayed close to 17%-19% of GDP regardless of tax rates, especially considering that none of these years are in the 1960-2000 range, which saw several huge tax changes and some severe economic events like the 1970's energy crisis. Even in the 2000's, more than half of the decade was in that range.
This seems to indicate that this is a very stable phenomena, regardless of the number of contributing factors. In fact, you can more strongly say that "Since WWII, Federal revenue has been 17%-20% of GDP regardless of essentially all factors, including tax rates" with a reasonable degree of accuracy.
Also, to say that over the 2000's it dropped from 20.6% to 14.9% is somewhat disingenuous, because it ignores the intermediate data, the rarity of 20% years, and the future predictions. It's more accurate to say that the 2000's started from an abnormally high 20.6% (probably in the wake of the dotcom bubble), slowly dipped to a low of 16.1% by 2004 (perhaps due to 9/11, the wars, and the Bush tax cuts), which was followed by a strong 17%-18% for four years, that was then followed by a crash to 14.9% in 2009 and 2010...which is expected to turn around by 2012 or 2013.
Machines making decisions for us isn't bad. However, there's two important conditions that need to be fulfilled for it to be a good thing: 1. The machines must make decisions on our behalf rather than just making decisions utilizing/involving us. In other words humanity needs to be "on top" instead of the machines "on top". I'm not saying that a robot manager or robot president that makes very high level decisions is a bad idea (and in fact I think it would probably be a good idea in the long run), but that such a robot needs to work for us (and not just working for a small number of human owner(s) either, otherwise the rest of us are screwed).
As an example, lets look at the coach robot. Beneficial: the owner purchases the robot and the team members use the robot to generate an effective training regimen and effective game day strategies, leading to more success for the team (and more profit for the owner) Dehumanizing: the owner purchases and appoints the robot, which drives the team members hard and controls them like puppets on the field, leading to more success for the team (and more profit for the owner)
The difference between these two scenarios is subtle and has more to do with the human-human (owner providing for team vs owner forcing team) and human-robot relationships (team uses robot vs team controlled by robot) than with the behavior of the robot (provides training regimen and strategy to team members in both cases). That said, there are scenarios where the robot's behavioral differences are critically important due to its incentives (ex. providing nigh-utopian environment for humans on their behalf vs eliminating most/all humans for resource optimization purposes).
2. Humans must maintain a liberal education long after it is useful in daily life. Critical thinking skills are essential to perform basic sanity checks.
This is an interesting thought, but there's a few additional issues.
1. Clearly this is not the only reason we need/want our own cars, several of which are covered by the other posters (privacy, customization, cars are a good place to store things you commonly take with you, etc.). Another such issue is that taxis are icky compared to one's own car (much like a public toilet compared to one's own toilet). Personal cars offer a better experience to people who can afford them. 2. You have to wait for the taxi to arrive after you order it. This can be particularly troublesome if you're not in a high traffic area, making the wait long. If you have your own car you don't have to think about this. It can also be an issue during high-load times as the taxis may not be able to get to you immediately if there's not enough of them (and there probably won't be enough as buying excessive taxis is unprofitable). While you can order ahead, you have to think about this beforehand and if you don't then a wait is inevitable. 3. This technology will probably require a human driver behind the wheel for some time into the future (for safety and/or legal reasons), eliminating the cost advantage until such a day that cars are frequently driver-less. At this point having your own car would be advantageous for other reasons (for instance, sending your car out to do chores on its own). Maybe eventually there'd be no difference, but it could be quite some time. 4. Cheap taxis would have little to no impact on personal vehicles with a use besides passenger transport, such as trucks. Though perhaps this could lead to variants on the taxi model that would lead to an impact on such vehicles, such as a "cargo taxi". 5. At least here in the US there's a car ownership culture. What kind of car you own is an important status symbol and that culture is unlikely to change in the face of driver automation. This will be especially true if taxis are very cheap.
On the other hand, taxis would indeed become a much better choice than they are now, displacing other forms of transport. In particular it would compete head to head with buses, which fill a similar niche. As you suggest, it would reduce the number of personal cars, but just how much is debatable.
Taxis do have certain advantages though: 1. Taking taxis everywhere could very well become substantially cheaper than operating a personal car. As you point out they make more journeys in a single day, spreading the operating costs over many people. 2. The parking situation is insane in certain areas of certain cities. Cheap taxis would be particularly popular in such areas (even amongst people who own a personal car). Though this would eventually lead to a balance between taxis and cars as the parking situation improves due to fewer people parking in the same area.
How much reduction in the number of cars would this create? Probably some (especially in the personal car market), but you could even see an *increase* in the total number of cars on produced. If the cost of taxis becomes comparable to buses then a massive number of taxis would replace the bus system. Unless most people also dropped their personal cars in favor of using taxis then there'd be more cars overall.
I don't think self-driving cars will use a different interior layout until most people deeply trust the automated driving system, which won't be for a long time (easily decades). Until then there'll be a human driver ready to take over, and in that case you need something similar to today's interior layout.
Also, while I'm sure larger automated personal vehicles that support walking around would exist, I think they will be too expensive to be common. A larger vehicle is simply more expensive than a smaller one and less fuel efficient/more maintenance heavy to boot. Most people will continue to use cheap but relatively cramped designs out of economic necessity (or a desire to save money for other more important things). Such larger designs make sense for buses or trains because many people share them, but what's the point in walking around in a vehicle with a SUV-sized floor containing just you and perhaps a handful of other people?
That said, perhaps designs that de-emphasize safety in exchange for comfort could appear. I could imagine seats that could be moved around to fit into other less safe but more comfortable or useful arrangements or beds becoming common (perhaps one could get a few more minutes of shuteye before work). Of course, the automated system would have to be fantastically reliable for people to have that level of trust...so I don't think this will be the case for a long time if at all.
No, the CLI appeals primarily to people who like to focus on memorising semantic minutiae and believe that doing so is, in and of itself, a productive endeavour.
Um, the semantic minutiae is an unfortunate obstacle that anyone who needs to use a CLI needs to overcome to benefit from it, not the point of the thing. Most CLI users memorize those things once and that's it. It's much like learning the semantic minutiae of calculus so that one may acquire a useful tool.
I use both GUI and CLI interfaces everyday. I tend to use GUIs more because most of the tools I use all the time are GUI-based and they're easy to use for human-centric tasks (web browsing, playing video, e-mail, document writing, etc.).
However, CLIs have several advantages that GUIs can't match (at least without mimicking or indirectly using CLIs). They're extremely easy to program for as the user interface is simple and extremely consistent...and most programming languages are set up to program for CLIs by default. Automation is much easier as you can naturally write scripts that use other programs, whereas a GUI-only interface would require a human to walk through those steps (or a program to blindly click through, making those nice pictures useless and error checking problematic). CLI commands can be very concise and powerful, saving time for the user who knows what they're doing compared to using a GUI. GUI programs have to be heavyweight with virtually every automation task included as a pre-programmed feature (and of course the developers won't be able to fit them all in so some users will lose out), whereas individual CLI programs can be small and do one relatively simple thing as they can be chained together. etc.
That said, CLIs are most advantageous to people who can program. If one can't program then their ability to automate via scripts and write simple but useful programs doesn't exist anyway, so the advantages of GUIs (low learning curve, beauty, human-centric design, etc.) far outweigh the seemingly esoteric advantages of CLIs. Plus, even programmers can benefit from well designed GUI programs, so it's not like it's just for the "masses".
However, CLIs do still have an advantages for the general public so they should not be knocked too hard. When programmers and other advanced users can increase their own productivity or even put together something that can benefit all users that know enough to type a simple command or install a plugin, then we all benefit. Learning the CLI is also often a first step toward becoming a programmer, and we need more of them. Therefore, we should maintain CLI functionality alongside GUI functionality as we move forward.
I think you're missing the point. The article is completely speculative and the video is basically about Microsoft's AI research. The guy in the video is talking about making systems that can adapt to the very problems you are talking about.
For instance, a dumb system will just give you an error when the filetype is wrong and stop there. An intelligent system on the other hand can decide that it needs to look at the file more deeply to determine what type it really is (using something like the Unix command "file") and then act on this information. We don't even need a particularly smart system to do something like this.
As for collection of unknown information needed for decision making, there's the same kinds of sources that humans use. Much pertinent information is available in standard formats online. The machine may have its own sensors to collect data. Oh, and of course the device can just ask the user, which is needed for details like which basic method the user plans to get to their destination.
Will MS try to lock us in? Sure. Will they exploit open sources of information while making their own sources impossible to use without their products? Probably. Will there be problems and mistakes? Of course. Is the idea of incorporating AI into a UI to make it easier to use ahead of its time? Maybe. Are these problems insurmountable? No. Maybe it won't be MS, but one day machines will be able to navigate most if not all of the hazards you've laid out...and in fact the idea laid out in the video is to attack these very problems head on.
There is nothing wrong with revolutionizing consumer applications, however, the article stated that it was revolutionizing robotics, not consumer applications.
First of all, your link just shows that altruism has an inherently selfish biological basis. Altruism is beneficial, therefore it is in our self interest to pursue it. Social structures that Libertarians love like capitalistic free markets embody this. "It is not from the benevolence of the butcher, the brewer, or the baker, that we can expect our dinner, but from their regard to their own interest."
Secondly, people being inherently selfish doesn't preclude altruistic behavior. It simply means that most "altruism" is really self-interested and real altruism is much rarer than selfish behavior and self-interested altruism.
Thirdly, what's wrong with building a system from a defensive position? Wouldn't a system designed for selfish people filled with altruistic people work out just fine? Any inequities in the system could be easily handled through charity and other altruistic practices. The same can't be said the other way around...an altruism-based system would fall apart in no time if filled with highly self-interested people...maybe even just a handful of them, as they'd be able to abuse everyone else's altruism.
Fourthly, if altruism is such a strong force...then why don't most people living in developed countries give a rat's ass about poverty, starvation, and disease outside of their own country? I have a "bleeding heart" friend who was suggesting we distribute the wealth in the US so that the poor can have a decent standard of living, but when I suggested that we should go further and distribute it worldwide he did not like that idea...I guess $10,500 a year isn't enough for him. Oh well, out of sight out of mind, amirite? Which is exactly the problem...we don't benefit personally from this kind of charity as we don't interact with the starving children in third world gutters, so most people are not interested in it beyond what it can do to make them look good to those around them...
Lastly, your comment about "rebuilding your world-view" is arrogant, condescending, and wholly premature. It seems to me that it will just put off others from listening to you seriously, dismissing you as a troll. Plus, your argument about altruism is weakened by the inclusion of childish jabs at your opponent that are just there to make you feel smarter than them.
What's wrong with revolutionizing consumer applications?
A Kinect may not be good enough for the applications you mention, but that doesn't mean they're insufficient for all applications. Who knows what kind of innovative low cost applications will emerge from this? I have a good friend in robotics that is doing his thesis using the Kinect and what he's working on has some interesting commercial potential (though of course I can't say any more without giving away my friend's idea).
Also, you say that improved future versions would "no longer be a kinect". While that may be strictly true, it's a lot like saying that future portable media players would no longer be an iPod (and therefore the iPod isn't revolutionary). Clearly it was revolutionary because it was the first one. The same applies to the Kinect. Maybe we'll generically call them robotic vision modules or something, but if the Kinect started the RVM trend then it will be just as revolutionary as the iPod.
It's revolutionary in the sense of "causing a revolution".
The ongoing Egyptian revolution has thus far consisted of popular protests and a bloodless military coup, both well worn political strategies. Yet these events are revolutionary for Egypt. Things will be different in Egypt in the future.
Wouldn't you say that people involved in sparking and carrying out the revolution were revolutionary? If a new technology came on the scene and caused the revolution, wouldn't it be a revolutionary technology?
In fact, isn't this the whole point of the word "revolutionary"? If a "revolutionary" new technology doesn't cause a revolution then was it really revolutionary at all? Maybe you're thinking of the word novel instead? You've made a good argument that the Kinect isn't novel.
In any case, if the Kinect is causing a robotics revolution then I'd say it's revolutionary regardless of the reason it is causing the revolution.
The good news is that it shows how much we can still improve microprocessor technology. Perhaps Moore's Law (or something similar) will keep up for quite some time into the future?
A better analogy would be that they are trying to build cars using steam engines. The core technology isn't quite there yet, but there's not much that chassis developers can do to speed that up. In the meanwhile, they'll try things out and maybe they'll find some use for what they do come up with (if we put these steam cars on rails...) and maybe they'll come up with some useful stuff that speeds up development when the core technology matures (we need better lubrication than tallow for all these mechanical parts).
Anyway, the main thing they're interested in here is perfecting robot bodies, not the AI that controls them. People will never take cute robots as seriously as other people and anything less than nigh-perfect human simulations will fall straight into the uncanny valley. Fixing these problems requires a very different skill-set than fixing the problems of NLP.
Perhaps it will turn out that this work is ahead of its time, but even so the lessons learned today can be applied in the future.
Amazingly, the US has not exploded into violence over a difference of opinion for over 140 years, and that one time it did happen there were critical economic factors that provided the bulk of the fuel for the conflagration.
I really don't see how the intellectual tensions between secularism and religion could lead to a civil war. Would enough Americans really be willing to fight and die over such ideas? I'm sure you can find some individuals on both sides who would be willing and able, but does the average person really care enough about this issue for it to get out of control?
While this is an interesting thought, I don't think it will be practical. Here's some thoughts on why: 1. It's more complex/expensive to assemble a customized car than a stock car. This could get quite costly. 2. It's more difficult to deal with testing and regulation.
Looking back on it now, I kinda messed up with the quote I picked. I meant general growth (economic, scientific, technological, etc.) not necessarily population growth. Why would we want to stop additional growth in these areas if we can hold onto our gains? Don't we want a bigger pie for when we eventually become more stable? That will mean there's more to divide up.
Some more population growth might even make sense in the context of expanding throughout the solar system or to largely uninhabited areas of the Earth or into newly minted virtual worlds, as we might not be able to fully utilize these resources with our current population, but beyond a certain point population growth means fewer resources per person and/or higher levels of inequality. The only advantages of population growth go to those people who don't exist yet. Since we exist already, we can be selfish and decide that all those other people don't need to exist so we can keep the extra resources to ourselves. It might not even be all that selfish if bringing them into the world would make their lives not worth living.
Also, at least for now we need to keep on reproducing. As long as people grow old and frail then if a nation (or the whole world) falls below the replacement rate, then it will mean there will not be enough young people to care for the old people and eventually we'll grow weak and vanish. Even for stability you need an average of a little over 2 children per female.
In fact, "growth" has become something of a religion itself. In public discourse and political debate, no one ever talks about stability; the need to "grow the economy" is taken as a "given", a commandment from on high. If a company's sales are merely stable from one quarter or year to the next, they are considered unsuccessful (or would be if the economy as a whole weren't currently shrinking).
For roughly the last 200 years, the economy has been growing exponentially (along with several other factors like population and computer speed during different periods). If a company's sales are merely stable for a period, then they are shrinking exponentially in comparison to the economy as a whole. It means that the company is losing relevance and customers are spending less and less of their incomes on the company's products (presumably because they are spending it on competing products). Nations with zero economic growth are similarly passed up by other economically growing nations, losing their relevance and strategic position. In this kind of environment it is unsurprising that stagnation is viewed as backwards.
In an overall stable system though, growth is unusual and exciting while stability is the norm. I'm sure people would get used to it quickly. After all, if a businessman or politician knows they can't deliver on growth, why would they hold themselves or others up to that lofty standard? It would make them look bad to say "growth" and deliver stagnation.
There will come a day when that trend stops, whether it's in 2030 or probably much later. The only question is whether we'll bring population growth to a "controlled landing" or to a crash.
How far can we grow without losing it all back when it stops is a very important question, perhaps even more relevant than whether the transition will be rough. Will our growth stop today? or will it end with us building a Dyson sphere and settling down into a cozy virtual world of our own design? or maybe somewhere in-between? If we still have a long way to grow, then we'd best not put the brakes on now.
Throughout most of the US, transport from airport to final destination relies on three basic methods: 1) Have a friend/relative/fellow employee pick you up in their car 2) Rent a car 3) Use paratransit or a taxi to get to where a car is available (or to a final destination where you don't plan on traveling far for the duration of the trip)
If decent intra-city public transportation exists at the location, it is a fourth option. Though this only works if your destination is the city containing the airport. There may be other options, but they are rarer.
I recently went to a family reunion in Colorado, we took a jitney to the hotel hosting the reunion and stayed there pretty much the whole time. We got to the handful of further away places we visited via local family members' cars.
Slashdot Mirror
You can already call into question any aspect of anything that's studied at any research institution. There's no need to acquire special justification. Questioning past results is an important part of science. Otherwise we'd have the Sun revolving around the Earth and supposed cold fusion.
That said, one should really have a logical argument based on all the relevant available facts when calling such things into question. Otherwise it's all just a waste of time for everyone involved...particularly the scientists involved, who could be making new discoveries in that time.
In this case, the best one can say is that the idea that 50 million refugees would materialize so soon was wrong, and that Dr. Bogardi of the UNU made a serious mistake in his prediction, which was possibly driven by the desire for creating scary headlines. That last bit about the headlines and anything further is speculation and politics, which is outside the realm of science.
The problem with expanding human rights is that in order for human rights to be meaningful, it must be possible to not trample upon them in any reasonable situation.
If you say that access to food is a human right, then what happens when a supervolcano erupting causes a massive food shortage due to long term lack of light? If we simply don't have enough food, how can we possibly fulfill everyone's right to eat? It comes down to providing everyone insufficient food to survive the disaster (leading to everyone dying) or keeping some people alive by abandoning others early on. Either way we're trampling this right, but the right thing to do is pro-actively trample on it right away so some survive. Would we be monsters for letting people die of starvation when we still had food?
How can we seriously consider something that may reasonably be impossible to fulfill a "right"?
Compare this to the right to free speech. It may be convenient or even beneficial to suppress this right in certain situations, but there's not a situation where it must be suppressed out of physical necessity. Even someone shouting "FIRE" in a crowded theater can be allowed to do so and even allowed to walk free afterwards...
Perhaps what is needed is a new word, or another old word, to describe the sort of "extended rights" that rely on a functioning economic system and the labor of others to provide. Entitlement seems to be popular, but it may fall into the same trap as "right" does...are we all entitled to food in the supervolcano scenario?
Maybe we should turn this on its head and put an emphasis on those providing the benefit to those in need of it? Instead of it being a right for the poor to eat or have access to the Internet, maybe we should see it as an obligation of those with the power to provide these things to do so...a sort of modern noblesse oblige.
Good thing it's coming from Taiwan, not China then.
While you are technically correct, looking at the data you linked to there are only 8 years over that 65 year time period where it differed from a 17%-20% average: 1949-1951, 1959, 2003-2004, and 2009-2010. If you increase the range to 16%-20% then it falls to a mere 4 years, all of which were in the 14% range: 1949-1950 and 2009-2010. Also, these years tend to be clustered together, which probably indicates they were caused by short term events.
It seems to me that these anomalous years don't shoot down the basic idea that Federal revenue has stayed close to 17%-19% of GDP regardless of tax rates, especially considering that none of these years are in the 1960-2000 range, which saw several huge tax changes and some severe economic events like the 1970's energy crisis. Even in the 2000's, more than half of the decade was in that range.
This seems to indicate that this is a very stable phenomena, regardless of the number of contributing factors. In fact, you can more strongly say that "Since WWII, Federal revenue has been 17%-20% of GDP regardless of essentially all factors, including tax rates" with a reasonable degree of accuracy.
Also, to say that over the 2000's it dropped from 20.6% to 14.9% is somewhat disingenuous, because it ignores the intermediate data, the rarity of 20% years, and the future predictions. It's more accurate to say that the 2000's started from an abnormally high 20.6% (probably in the wake of the dotcom bubble), slowly dipped to a low of 16.1% by 2004 (perhaps due to 9/11, the wars, and the Bush tax cuts), which was followed by a strong 17%-18% for four years, that was then followed by a crash to 14.9% in 2009 and 2010...which is expected to turn around by 2012 or 2013.
Machines making decisions for us isn't bad. However, there's two important conditions that need to be fulfilled for it to be a good thing:
1. The machines must make decisions on our behalf rather than just making decisions utilizing/involving us. In other words humanity needs to be "on top" instead of the machines "on top". I'm not saying that a robot manager or robot president that makes very high level decisions is a bad idea (and in fact I think it would probably be a good idea in the long run), but that such a robot needs to work for us (and not just working for a small number of human owner(s) either, otherwise the rest of us are screwed).
As an example, lets look at the coach robot.
Beneficial: the owner purchases the robot and the team members use the robot to generate an effective training regimen and effective game day strategies, leading to more success for the team (and more profit for the owner)
Dehumanizing: the owner purchases and appoints the robot, which drives the team members hard and controls them like puppets on the field, leading to more success for the team (and more profit for the owner)
The difference between these two scenarios is subtle and has more to do with the human-human (owner providing for team vs owner forcing team) and human-robot relationships (team uses robot vs team controlled by robot) than with the behavior of the robot (provides training regimen and strategy to team members in both cases). That said, there are scenarios where the robot's behavioral differences are critically important due to its incentives (ex. providing nigh-utopian environment for humans on their behalf vs eliminating most/all humans for resource optimization purposes).
2. Humans must maintain a liberal education long after it is useful in daily life. Critical thinking skills are essential to perform basic sanity checks.
This is an interesting thought, but there's a few additional issues.
1. Clearly this is not the only reason we need/want our own cars, several of which are covered by the other posters (privacy, customization, cars are a good place to store things you commonly take with you, etc.). Another such issue is that taxis are icky compared to one's own car (much like a public toilet compared to one's own toilet). Personal cars offer a better experience to people who can afford them.
2. You have to wait for the taxi to arrive after you order it. This can be particularly troublesome if you're not in a high traffic area, making the wait long. If you have your own car you don't have to think about this. It can also be an issue during high-load times as the taxis may not be able to get to you immediately if there's not enough of them (and there probably won't be enough as buying excessive taxis is unprofitable). While you can order ahead, you have to think about this beforehand and if you don't then a wait is inevitable.
3. This technology will probably require a human driver behind the wheel for some time into the future (for safety and/or legal reasons), eliminating the cost advantage until such a day that cars are frequently driver-less. At this point having your own car would be advantageous for other reasons (for instance, sending your car out to do chores on its own). Maybe eventually there'd be no difference, but it could be quite some time.
4. Cheap taxis would have little to no impact on personal vehicles with a use besides passenger transport, such as trucks. Though perhaps this could lead to variants on the taxi model that would lead to an impact on such vehicles, such as a "cargo taxi".
5. At least here in the US there's a car ownership culture. What kind of car you own is an important status symbol and that culture is unlikely to change in the face of driver automation. This will be especially true if taxis are very cheap.
On the other hand, taxis would indeed become a much better choice than they are now, displacing other forms of transport. In particular it would compete head to head with buses, which fill a similar niche. As you suggest, it would reduce the number of personal cars, but just how much is debatable.
Taxis do have certain advantages though:
1. Taking taxis everywhere could very well become substantially cheaper than operating a personal car. As you point out they make more journeys in a single day, spreading the operating costs over many people.
2. The parking situation is insane in certain areas of certain cities. Cheap taxis would be particularly popular in such areas (even amongst people who own a personal car). Though this would eventually lead to a balance between taxis and cars as the parking situation improves due to fewer people parking in the same area.
How much reduction in the number of cars would this create? Probably some (especially in the personal car market), but you could even see an *increase* in the total number of cars on produced. If the cost of taxis becomes comparable to buses then a massive number of taxis would replace the bus system. Unless most people also dropped their personal cars in favor of using taxis then there'd be more cars overall.
I don't think self-driving cars will use a different interior layout until most people deeply trust the automated driving system, which won't be for a long time (easily decades). Until then there'll be a human driver ready to take over, and in that case you need something similar to today's interior layout.
Also, while I'm sure larger automated personal vehicles that support walking around would exist, I think they will be too expensive to be common. A larger vehicle is simply more expensive than a smaller one and less fuel efficient/more maintenance heavy to boot. Most people will continue to use cheap but relatively cramped designs out of economic necessity (or a desire to save money for other more important things). Such larger designs make sense for buses or trains because many people share them, but what's the point in walking around in a vehicle with a SUV-sized floor containing just you and perhaps a handful of other people?
That said, perhaps designs that de-emphasize safety in exchange for comfort could appear. I could imagine seats that could be moved around to fit into other less safe but more comfortable or useful arrangements or beds becoming common (perhaps one could get a few more minutes of shuteye before work). Of course, the automated system would have to be fantastically reliable for people to have that level of trust...so I don't think this will be the case for a long time if at all.
That's because Star Control II is available for free now: http://sc2.sourceforge.net/
No, the CLI appeals primarily to people who like to focus on memorising semantic minutiae and believe that doing so is, in and of itself, a productive endeavour.
Um, the semantic minutiae is an unfortunate obstacle that anyone who needs to use a CLI needs to overcome to benefit from it, not the point of the thing. Most CLI users memorize those things once and that's it. It's much like learning the semantic minutiae of calculus so that one may acquire a useful tool.
I use both GUI and CLI interfaces everyday. I tend to use GUIs more because most of the tools I use all the time are GUI-based and they're easy to use for human-centric tasks (web browsing, playing video, e-mail, document writing, etc.).
However, CLIs have several advantages that GUIs can't match (at least without mimicking or indirectly using CLIs). They're extremely easy to program for as the user interface is simple and extremely consistent...and most programming languages are set up to program for CLIs by default. Automation is much easier as you can naturally write scripts that use other programs, whereas a GUI-only interface would require a human to walk through those steps (or a program to blindly click through, making those nice pictures useless and error checking problematic). CLI commands can be very concise and powerful, saving time for the user who knows what they're doing compared to using a GUI. GUI programs have to be heavyweight with virtually every automation task included as a pre-programmed feature (and of course the developers won't be able to fit them all in so some users will lose out), whereas individual CLI programs can be small and do one relatively simple thing as they can be chained together. etc.
That said, CLIs are most advantageous to people who can program. If one can't program then their ability to automate via scripts and write simple but useful programs doesn't exist anyway, so the advantages of GUIs (low learning curve, beauty, human-centric design, etc.) far outweigh the seemingly esoteric advantages of CLIs. Plus, even programmers can benefit from well designed GUI programs, so it's not like it's just for the "masses".
However, CLIs do still have an advantages for the general public so they should not be knocked too hard. When programmers and other advanced users can increase their own productivity or even put together something that can benefit all users that know enough to type a simple command or install a plugin, then we all benefit. Learning the CLI is also often a first step toward becoming a programmer, and we need more of them. Therefore, we should maintain CLI functionality alongside GUI functionality as we move forward.
I think you're missing the point. The article is completely speculative and the video is basically about Microsoft's AI research. The guy in the video is talking about making systems that can adapt to the very problems you are talking about.
For instance, a dumb system will just give you an error when the filetype is wrong and stop there. An intelligent system on the other hand can decide that it needs to look at the file more deeply to determine what type it really is (using something like the Unix command "file") and then act on this information. We don't even need a particularly smart system to do something like this.
As for collection of unknown information needed for decision making, there's the same kinds of sources that humans use. Much pertinent information is available in standard formats online. The machine may have its own sensors to collect data. Oh, and of course the device can just ask the user, which is needed for details like which basic method the user plans to get to their destination.
Will MS try to lock us in? Sure. Will they exploit open sources of information while making their own sources impossible to use without their products? Probably. Will there be problems and mistakes? Of course. Is the idea of incorporating AI into a UI to make it easier to use ahead of its time? Maybe. Are these problems insurmountable? No. Maybe it won't be MS, but one day machines will be able to navigate most if not all of the hazards you've laid out...and in fact the idea laid out in the video is to attack these very problems head on.
There is nothing wrong with revolutionizing consumer applications, however, the article stated that it was revolutionizing robotics, not consumer applications.
Make that consumer applications...of robotics.
Uh, really?
First of all, your link just shows that altruism has an inherently selfish biological basis. Altruism is beneficial, therefore it is in our self interest to pursue it. Social structures that Libertarians love like capitalistic free markets embody this. "It is not from the benevolence of the butcher, the brewer, or the baker, that we can expect our dinner, but from their regard to their own interest."
Secondly, people being inherently selfish doesn't preclude altruistic behavior. It simply means that most "altruism" is really self-interested and real altruism is much rarer than selfish behavior and self-interested altruism.
Thirdly, what's wrong with building a system from a defensive position? Wouldn't a system designed for selfish people filled with altruistic people work out just fine? Any inequities in the system could be easily handled through charity and other altruistic practices. The same can't be said the other way around...an altruism-based system would fall apart in no time if filled with highly self-interested people...maybe even just a handful of them, as they'd be able to abuse everyone else's altruism.
Fourthly, if altruism is such a strong force...then why don't most people living in developed countries give a rat's ass about poverty, starvation, and disease outside of their own country? I have a "bleeding heart" friend who was suggesting we distribute the wealth in the US so that the poor can have a decent standard of living, but when I suggested that we should go further and distribute it worldwide he did not like that idea...I guess $10,500 a year isn't enough for him. Oh well, out of sight out of mind, amirite? Which is exactly the problem...we don't benefit personally from this kind of charity as we don't interact with the starving children in third world gutters, so most people are not interested in it beyond what it can do to make them look good to those around them...
Lastly, your comment about "rebuilding your world-view" is arrogant, condescending, and wholly premature. It seems to me that it will just put off others from listening to you seriously, dismissing you as a troll. Plus, your argument about altruism is weakened by the inclusion of childish jabs at your opponent that are just there to make you feel smarter than them.
What's wrong with revolutionizing consumer applications?
A Kinect may not be good enough for the applications you mention, but that doesn't mean they're insufficient for all applications. Who knows what kind of innovative low cost applications will emerge from this? I have a good friend in robotics that is doing his thesis using the Kinect and what he's working on has some interesting commercial potential (though of course I can't say any more without giving away my friend's idea).
Also, you say that improved future versions would "no longer be a kinect". While that may be strictly true, it's a lot like saying that future portable media players would no longer be an iPod (and therefore the iPod isn't revolutionary). Clearly it was revolutionary because it was the first one. The same applies to the Kinect. Maybe we'll generically call them robotic vision modules or something, but if the Kinect started the RVM trend then it will be just as revolutionary as the iPod.
It's revolutionary in the sense of "causing a revolution".
The ongoing Egyptian revolution has thus far consisted of popular protests and a bloodless military coup, both well worn political strategies. Yet these events are revolutionary for Egypt. Things will be different in Egypt in the future.
Wouldn't you say that people involved in sparking and carrying out the revolution were revolutionary? If a new technology came on the scene and caused the revolution, wouldn't it be a revolutionary technology?
In fact, isn't this the whole point of the word "revolutionary"? If a "revolutionary" new technology doesn't cause a revolution then was it really revolutionary at all? Maybe you're thinking of the word novel instead? You've made a good argument that the Kinect isn't novel.
In any case, if the Kinect is causing a robotics revolution then I'd say it's revolutionary regardless of the reason it is causing the revolution.
Really that was just really bad. Satellites have never been "limited" to the size and weight of Hydrogen bombs.
Except that's not really what he said...the very next paragraph elaborates on this point.
Even if the leak has no negative impact on the effectiveness of the leaked policies, it is still ironic.
It also clearly demonstrates why such policies are being developed.
The good news is that it shows how much we can still improve microprocessor technology. Perhaps Moore's Law (or something similar) will keep up for quite some time into the future?
A better analogy would be that they are trying to build cars using steam engines. The core technology isn't quite there yet, but there's not much that chassis developers can do to speed that up. In the meanwhile, they'll try things out and maybe they'll find some use for what they do come up with (if we put these steam cars on rails...) and maybe they'll come up with some useful stuff that speeds up development when the core technology matures (we need better lubrication than tallow for all these mechanical parts).
Anyway, the main thing they're interested in here is perfecting robot bodies, not the AI that controls them. People will never take cute robots as seriously as other people and anything less than nigh-perfect human simulations will fall straight into the uncanny valley. Fixing these problems requires a very different skill-set than fixing the problems of NLP.
Perhaps it will turn out that this work is ahead of its time, but even so the lessons learned today can be applied in the future.
Amazingly, the US has not exploded into violence over a difference of opinion for over 140 years, and that one time it did happen there were critical economic factors that provided the bulk of the fuel for the conflagration.
I really don't see how the intellectual tensions between secularism and religion could lead to a civil war. Would enough Americans really be willing to fight and die over such ideas? I'm sure you can find some individuals on both sides who would be willing and able, but does the average person really care enough about this issue for it to get out of control?
How can we know for sure that someone hasn't done this already?
While this is an interesting thought, I don't think it will be practical. Here's some thoughts on why:
1. It's more complex/expensive to assemble a customized car than a stock car. This could get quite costly.
2. It's more difficult to deal with testing and regulation.
Looking back on it now, I kinda messed up with the quote I picked. I meant general growth (economic, scientific, technological, etc.) not necessarily population growth. Why would we want to stop additional growth in these areas if we can hold onto our gains? Don't we want a bigger pie for when we eventually become more stable? That will mean there's more to divide up.
Some more population growth might even make sense in the context of expanding throughout the solar system or to largely uninhabited areas of the Earth or into newly minted virtual worlds, as we might not be able to fully utilize these resources with our current population, but beyond a certain point population growth means fewer resources per person and/or higher levels of inequality. The only advantages of population growth go to those people who don't exist yet. Since we exist already, we can be selfish and decide that all those other people don't need to exist so we can keep the extra resources to ourselves. It might not even be all that selfish if bringing them into the world would make their lives not worth living.
Also, at least for now we need to keep on reproducing. As long as people grow old and frail then if a nation (or the whole world) falls below the replacement rate, then it will mean there will not be enough young people to care for the old people and eventually we'll grow weak and vanish. Even for stability you need an average of a little over 2 children per female.
In fact, "growth" has become something of a religion itself. In public discourse and political debate, no one ever talks about stability; the need to "grow the economy" is taken as a "given", a commandment from on high. If a company's sales are merely stable from one quarter or year to the next, they are considered unsuccessful (or would be if the economy as a whole weren't currently shrinking).
For roughly the last 200 years, the economy has been growing exponentially (along with several other factors like population and computer speed during different periods). If a company's sales are merely stable for a period, then they are shrinking exponentially in comparison to the economy as a whole. It means that the company is losing relevance and customers are spending less and less of their incomes on the company's products (presumably because they are spending it on competing products). Nations with zero economic growth are similarly passed up by other economically growing nations, losing their relevance and strategic position. In this kind of environment it is unsurprising that stagnation is viewed as backwards.
In an overall stable system though, growth is unusual and exciting while stability is the norm. I'm sure people would get used to it quickly. After all, if a businessman or politician knows they can't deliver on growth, why would they hold themselves or others up to that lofty standard? It would make them look bad to say "growth" and deliver stagnation.
There will come a day when that trend stops, whether it's in 2030 or probably much later. The only question is whether we'll bring population growth to a "controlled landing" or to a crash.
How far can we grow without losing it all back when it stops is a very important question, perhaps even more relevant than whether the transition will be rough. Will our growth stop today? or will it end with us building a Dyson sphere and settling down into a cozy virtual world of our own design? or maybe somewhere in-between? If we still have a long way to grow, then we'd best not put the brakes on now.
For the same reason that pointing out that a study was funded by an oil corporation gets modded up?
Throughout most of the US, transport from airport to final destination relies on three basic methods:
1) Have a friend/relative/fellow employee pick you up in their car
2) Rent a car
3) Use paratransit or a taxi to get to where a car is available (or to a final destination where you don't plan on traveling far for the duration of the trip)
If decent intra-city public transportation exists at the location, it is a fourth option. Though this only works if your destination is the city containing the airport. There may be other options, but they are rarer.
I recently went to a family reunion in Colorado, we took a jitney to the hotel hosting the reunion and stayed there pretty much the whole time. We got to the handful of further away places we visited via local family members' cars.