Just build more receivers. Certainly a power cable on the ocean floor circling the globe will be easier to build than a giant power plant on the Moon and a network of satellites, ground stations, receiver dishes, etc.
If the power will be indeed 20% of the noontime sunlight, won't it be less efficient to collect that energy than to collect the energy from the sunlight that already falls on the same receiving antenna? And get 5 times more power output at noon without any moon station and satellites?
Extending this to be able to work with every widget of a given GUI would not be very difficult, and the only difference between that and a themed GUI is instead of hardcoding the patterns, you just load them.
Except that most of pattern appear at random times (you can't be sure that the piece of interface stopped changing at some point -- you can only guess that if it's not changing anymore for some time, your next input will be sent to it, and not to, say a dialog box that is about to appear). And multiple items may overlap, require scrolling, be obscured, etc. This is ambiguous and therefore unsafe to rely on.
How is it that a sequence of bits representing pixels is any different than a sequence of bits representing letters and words?
Pixels are not in a sequence on the screen -- they should be processed as a single image of the screen, and sequence is formed by a changing image. It's not even possible to get ALL changes that happened over some time, GUIs are not designed for that, and should be readable for a human even if multiple changes happened between screen refreshes. Obviously, there is always some guesswork about what really got obscured, users often find that they have to "wiggle" windows around just to find if their scrollbars are really where it seems, or if there is a hidden window under them.
On the other hand, text is always a continuous stream, and it does not matter at which precise moment some piece of it is being sent -- the semantics don't change with any buffering as long as everything that is sent is eventually received.
I'm not sure how much programming you've done
More than you likely ever will.
, but you need to do more before you can authoratively say that something can't be done, especially when I can come along and say that I have done it.
Game interfaces are borderline cases that do not correspond to any traditional GUI guidelines, therefore solutions for them do not matter.
Writing a program to understand a given GUI would be fairly trivial with decent OCR and graphics libraries.
False.
Writing a program to understand a themable GUI with access to the themes and knowledge of what theme is being used would be harder, but not impossible.
False
Writing a program to understand any GUI understandable by humans would be... well impossible if you're talking about skinned MP3 players and stuff, but even a marginally consistent GUI wouldn't be impossible, just VERY difficult.
False. All those things are impossible. After a fair amount of technology development it may become possible to do them with low probability of errors, but it will be always impossible to avoid errors completely because GUIs are inherently ambiguous. This is tolerable with humans at their terminals because ambiguity of GUIs is often outweighed by lower speed and higher probability of human errors without GUIs, however the problem remains with humans, too.
Computers do not have a preference for text. We could write a command shell that only accepted numbers, or only accepted certain tones played into a microphone, or frequencies of light, or weights, or densities, or whatever we could imagine.
All those things ARE different kinds of text, defined for the computer as sequences of bytesthat directly correspond to the data being sent across the interface.
If computers have a preference for anything, it's opened and closed circuts, and that's only if you say that computers must be a digital device.
Computers have preference for circuits, opened and closed IN SEQUENCE as means of their i/o. What is the same as sequentially written text.
Actually with a formalized legal system it would be extremely hard to convict ANYONE fro murder -- the amount and certainty of evidence will have to be far beyond what was present in OJ case. Howewer I am not talking about court proceedings that involve more hand-waving and psychological presure than anything that has to do with laws. I am talking about laws themselves and their interpretations that apply to life long before someone steps into court, plain questions about whatis legal and what is not.
Ex: anything that has to do with fair use of copyrighted material.
There's quite an oppurtunity cost for the computer illiterate to learn to program... let alone program well. It makes more sense for the computer illiterate individual to trade resources for computer software/services.
There is an even higher opportunity cost for being born, yet all people that you have ever seen, were born at some point.
People _do_ adapt to a toaster. They place bread in it, press on the lever and wait for toast to be ready.
A person that is not adapted to the toaster would demand that a toaster will emulate a pastry chef, take orders in spoken English, and give out ready toasts without burdening the user with bringing and slicing bread, placing it into the correct position, pressing levers or listening to a loud "Twang!!!" when the toast is ready.
Same about computers. People who advocate making computers adapt to humans want a computer to behave like a secretary, like a VCR, like a mathematician, physicist, engineer -- whatever but behaving like a computer. What is fine in some cases and completely wrong in others.
Computers have no preference for text. They have no preference for graphics. If they could be said to have any sort of preference at all, it would be binary. And that would still be a misleading statement.
Sure they do. Try to write a program that can operate on another computer, reading the bitmap ot its screen in GUI and sending keyboard/mouse events. Can't be done, there will be always ambiguous and hard to recognize situations (like, oddly overlapping windows and shapes, unfamiliar fonts, etc.), unless there is some severe limitation imposed on the GUI.
On the other hand, countless scripts talk to the terminals or other programs through text streams.
GUI is something that a computer program can not understand -- humans and only humans have the set of metaphors that allow them to see objects in what GUI draws on the screen. This set of metaphors is completely foreign to the computer itself, and its idea of objects is merely represented in the GUI picture.
They are only different because it is profitable for some companies to artificially limit the number of people who can do their jobs. Really there are only:
1. Data entry. 2. Secretary-like job of writing pretty letters. 3. Data processing. 4. Writing applications. 5. Writing tools for writing applications. 6. Writing infrastructure. 7. Writing tools for writing infrastructure.
First two can go feed some sharks, and I won't notice. Last four are busy enough writing code already. The third one is actually what that the question is about.
The idea that "data processing" is an uncreative job, and all such a person should do is taking pretty letters written by secretaries, demand from application programmers to write something that process them and outputs more pretty letters, and then using those applications like a good peon until there is again a reason to bitch at the application programmer, is widespread but ridiculous and insulting.
People who understand what data they deal with, should be at least able to express their procedures in a way that doesn't require a trip to the application programmer every time they need to summarize another column, or attach another database. And when those people are incapable of doing that because they were taught an extremely limited idea of how they are supposed to use a computer, the efficiency of their work drops to a ridiculous level, reducing them to the full time "data entry clerks".
Subsistence farming was extremely rare through the history, usually "peasant" was not just producing the food for his own family but had various kinds of obligation to "lord", "owner", "landowner", "government" that taxed him, controlled his behavior, and often property, and performed various services for him that the peasant was incapable of, lacking time, skills or education.
In other words, someone in a subservient position that requires him to perform large amount of agricultural work but prevents him from doing anything more complex.
GUI is useful. The point is, if ALL you know is GUI, you can't do anything that others didn't do for you, and you need to know more than that to do something less trivial. Sure, Photoshop (or Gimp) GUI works fine, but each and every piece of those GUI was made for a purpose that was already known at the time of its writing, and there is more to image editing than what is written and neatly packaged in those programs and plugins to them. It's a tool, and there is nothing wrong with tools, however inability to modify one's tools is always a limitation on the person's creativity, and at some point the user is faced with a situation where the tool is inadequate. This is where ignorant user has to do thousands of drags and drops, or has to train his aim to hit a particular pixel on the screen with a mouse, and knowledgeable one types few tens of characters.
There always will be something that "peasants" won't care to learn, and among that will be "literacy", something that truly differentiates them from more advanced parts of the society. Modern "peasants" may not be defined as people who do farming, and are placed under the rule of their lords, but there certainly there is a part of the society that survives while getting inadequate education, and having ridiculously low amount of decisions that they can make for themselves.
If something will improve this situation, it will be the elimination of the things that cause large number of people to become "peasants", however I am not holding my breath for that. In particular, modern American society's economy, politics and culture are entirely based on the presence of a large uneducated, faceless, easy to manipulate mass right above the bottom of the society and below the much smaller number of well-educated people.
Eventually when more complex knowledge will become a thing that separates peasants and lords, some of the current "lords" may find their heirs being thrown to the "peasants" level, and some "peasants" may find themselves "accidentally" acquiring literacy as a side effect of job training or as a result of their curiosity, but it's unlikely that the structure will get fundamentally changed because of that. It can only be changed if there will be a force that purposefully acts to eliminate the "illiteracy", and therefore the existence of "peasants" and their servitude. Too bad, American society has no such force.
There is one area of everyday human activity where complex logical structures are always expressed using natural language without any aid from more formalized languages and systems of expression.
It's called legal system, and it's a mess so huge, no one, human or machine, can predict its behavior, and an army of lawyers, judges and lawmakers of all kinds spend untold amounts of time trying to implement it in their various ways. If computer programming will ever turn into something that resembles this, determining which color should be a pixel (15, 351) ten seconds into a Quake game will be a process not unlike the Kobe Bryant case.
If everything is a stream of bytes, any kind of interface can be created with it, based on its purpose. Simple interfaces can hide complex structure by just exchanging blocks of data blindly, complex ones can use parsers (a thing that Windows programmers never learn) and implement any format easily. And this clearly separates the programs -- they never ever touch each other's address space and can have any, simple or complex, design of their objects. Also whatever can be done with pipes, can be done with sockets, thus allowing remote and distributed processing without any change to the software doing it.
Dotnet's "command line", judging by the descriptions that Microsoft made, allows things to just sit in one happy clusterfuck and mess with objects while they are freely floating inside the shell application, something that is in no way different from writing "scripts" in a large spreadsheet. This is yet another example how Microsoft invents 65537th iteration of the mix of their original DLL and DDE, two ideas that they still can't get right after more than a decade of development.
And not a single word about source and messages' leak that demonstrated how bad the system really is. Instead we see claims that Diebold has "empirical data" that voters love the new machines just like Happy Meals at Mcdonalds.
Microsoft confusion machine again
on
Microsoft's new CLI
·
· Score: 2, Interesting
So basically Microsoft, in their fine tradition of creating an confusion in terms, made a "shell" that can't run programs but instead runs "managed code" (aka shared libraries in dotnet environment) and maintains all objects inside itself, but has a syntax that is confusingly similar to Unix shell pipes. The concept is obviously flawed -- Unix shell uses Unix-specific unified file descriptor model, that Windows lacks, so while in Unix shell is often used to process easily-parseable text using pipelines of programs (some remote) and any overcomplication of data passed between programs is frowned upon, in Windows the most convoluted ever format is being used to exchange things between "programs" that are sitting in the belly of that "shell". They could just as well make the whole interface an UML editor.
If anything, it can prevent some people who learned that shit from switching to Unix -- they will until the end of their miserable lives associate pipelines with shitloads of DLLs instead of streams of text.
This is just barking up the wrong tree. Of course, there is a decline, but mesauring the accomplishment by the amount of standing out individuals is a wrong way to measure or find a decline. In general the amount of "celebrity cult" of scientists, politicians and writers through history decreases, as they become more numerous, and their accomplishments require more co-operation. In politics it can be seen when one compares absolute monarchies of the past when almost all political acts are attributed to a single person, to modern systems with large governments. However the same can be said about science, except that in a modern world when a group of politician starts a war, most of their names usually get known and mentioned, but when a group of scientists develops a powerful weapon, most of names are not mentioned, and certainly nothing is known about people who re-done the same work elsewhere.
Also different cultures have differrent traditions of crediting individuals to begin with, not to mention that large amount of material is published in the languages other than English/German/French/Spanish/Italian, and therefore is not being referenced by European sources, so the analysis of the "accomplishment" distribution through the world is even more skewed than the distribution of them through the time.
So the actual amount of progress beind made can't be measured by just individuals that stand out enough to get credited, at some point more accomplishment leads to less visible "leaders". However the decline in overall progress started in the second half of the 20th century is definitely intuitively visible. In my opinion, the problem is unrelated to personal achievements, and is based on the skewed and distorted structure of the global economy that formed at the same time. Basically, industrial revolution was based on the idea that if mechanisms won't increase people's productivity, nothing else will. When it became easier to "let someone else do this", the incentive to develop similar kinds of advances in science and engineering disappeared. As an example I can see it as great dishonor to humankind that my Levi's 501 jeans that I am wearing (and went through at least 10 pairs over my life so far) are not being produced on a humanless automated line, employing CNC cutting and sweing machines exclusively. It would need a pretty small amount of innovation in engineering, and maybe some work in science (machine vision that would be able to handle fabric texture properly, see threads and seams at odd angles, etc.). But it's not being done because sweatshops are cheaper, even if they are a horrendous waste of human effort otherwise.
The progress in technology became strongly skewed toward entertainment-related purposes because they provide higher return on investment compared to similar effort made "sweatshop-style", however I doubt that there is a natural process that can extend those advances into other areas until the global economy stabilizes to make domestic industry a necessity for all countries again. That would basically mean a return of industriak revolution, though likely under other name, and with a lot of pundits claiming that it's something "even better than our postindustrial world". In my opinion, "postindustrial" societies are merely declined elite, fed by sweatshops, not unlike arsitocracy/peasants-based feudal societies that preceded the industrial revolution in the first place.
Thought recognition technology has actually gone a lot further than what you imply (I first saw a report on this sort of thing 2-3 years back).
This is barely any recognition at all -- it uses signals from few neurons, maps them to movements parameters, and allows visual feedback. Feedback also simplifies the procedure by allowing the brain to "re-learn" how it is supposed to control the movement, as opposed to device learning how brain does that already.
It can be called "thought" only because it's not a signal collected from nerves going to the muscles, but its predecessor, however it contains the same amount and kind of information. No symbolic or abstract information is passed, so it amounts to giving the user an equivalent of joystick or mouse, that is harder to control than an actual joystick or mouse, a thing that never was a problem for a wearable computer in the first place. Again, beyond that level the problem is fundamental, and would require multiple levels of recognition, each involving reverse-engineering a mechanism that we know almost nothing about.
Even if you regard the above as "most basic functions", that functionality is more than good enough for what I'm talking about. Of course the current technology is quite invasive and drastic. So I don't see that being very popular anytime soon.
Also it's still impossible to get this signal without damaging large number of neurons in the process of installing the sensors, and the lifetime of electrode is limited.
Now if they get to the stage of giving people SVGA/s-video inputs, things could get even more interesting.
Actually there were some successful tests of prosthetic vision devices. Just as invasive, produces a poor resemblance of an image, and involves re-learning of the image processing in the brain to whatever mapping that the device produces, as opposed to imitating the mapping originally used to process signals from the eyes (that, again, is very little known about beyond the basic encoding principles that the device imitates).
Two cameras (one looking forward, another looking at the pupil), and one in-glasses projection display per eye will provide the same functionality better, and a chord keyboard, or twiddler-like single-hand keyboard already can provide good enough symbolic input, at least until some fundamental discovery of the processing/reverse engineering procedures for thought mechanisms will be made.
It is not. If handwritten text recognition from a photograph can't be done, an part of your device's purpose is lost -- you can only manually convert those photographs to text, and unless you are in a hurry when taking those, and have plenty of time to convert them to text, you gain nothing compared to just taking notes. It will be an equivalent of using an audio record of a lecture to make notes later -- often a good compromise but certainly not a replacemnt for instant recording. And since there are fundamental problem with this kind of "handwriting recognition", the solution isn't going to be available any time soon.
The rest can be added later - e.g. thought pattern recognition (already done with rats, apes and monkeys),
Those only could "recognize" the most basic functions, at the expense of very intrusive sensors and large measurement equipment. Again, there are fundamental problems with higher levels of functionality, and no noticeable progress on those was made in at least 20 years, placing this work in the same category as fusion energy. So not any time soon, either.
face recognition (this doesn't have to be 100% accurate - coz you can quickly verify it's your friend/enemy/customer/boss) and so on.
Face recognition is so far from those tasks, ans can be reduced to such a simple statistical manipulations on geometry, it's primitive compared to those tasks. The fact that it's still unreliable only demonstrated how far we are from the solution for the rest of those problems.
There's a reason for that, and it's tied into something that a lot of people here on slashdot and similar tech-based sites don't understand: most people don't want to have to learn something new just to get their job done.
If they don't have sufficient knowledge to do their work, doesn't it mean that they are incompetent? Sure, if there are alternatives that don't require such a knowledge while still letting them do their job, no (or little) harm is done, but so far most of the attempts to make an executive's work easier to perform without requiring any new knowledge, failed miserably. The nature of work that involves dealing with large amounts of text, numbers and various representations of those, absolutely demands an ability to read and write them.
Substituting text with "digital ink" that is an inferior substitute of text (can't be processed, hard to read), and numbers with the same "digital ink" that is a completely inadequate replacement for numbers (recognition errors can't be easily found and fixed), leaves people who use those just as incompetent as if they used pen and paper exclusively.
Executives, as you point out, don't need to type because they "have someone to do that for them".
If this is true, and an executive is capable of doing his work with all text input being done by secretary, he gains nothing from replacing a secretary with a pad that he writes squiggles on. Usually the real problem is that the job demands closer interaction with the data, and a secretary can't do all that efficiently. But then the solution offered should be more efficient than a secretary, not less efficient like the tablet design is -- then it's better to just keep a secretary.
A lecture hall of 250 students waiting a whole minute for acrobat reader to load is not fast enough.
The loading time of Acrobat Reader has little to do with the CPU speed, and A LOT with the use of a slow notebook hard drive (in this case even designed to work when being rapidly tilted), and RAM+swap being filled with crap.
- handwriting recognition that takes pen movement as an input, and allows user to correct the mis-recognized text (this is what everything uses -- with moderate success),
- handwriting recognition that uses pen movement as an input without a feedback (what can work, but usually only after either user or computer will go through a lot of learning in a feedback mode),
- "handwriting recognition" that takes the images of written text as an input (what is something that even humans often have problems with, and for the modern recognition technology is still in a "don't even try" area).
Actually you will make a potent TARGET -- the probability of it being hit nears 1 in any imaginable battle because no one among the enemy will shoot at anything else until it will be down.
If you think about it, the only way to "abolish war" is peace through superiority. Nobody will directly confront an obviously superior opponent.
I would advise the military of all nuclear-capable countries, other than US, to once in a year nuke a city where the person, who made the most recent statement of this kind, resides, until everyone gets the idea. You know, to show how wrong they are.
Slashdot Mirror
Just build more receivers. Certainly a power cable on the ocean floor circling the globe will be easier to build than a giant power plant on the Moon and a network of satellites, ground stations, receiver dishes, etc.
If the power will be indeed 20% of the noontime sunlight, won't it be less efficient to collect that energy than to collect the energy from the sunlight that already falls on the same receiving antenna? And get 5 times more power output at noon without any moon station and satellites?
Extending this to be able to work with every widget of a given GUI would not be very difficult, and the only difference between that and a themed GUI is instead of hardcoding the patterns, you just load them.
Except that most of pattern appear at random times (you can't be sure that the piece of interface stopped changing at some point -- you can only guess that if it's not changing anymore for some time, your next input will be sent to it, and not to, say a dialog box that is about to appear). And multiple items may overlap, require scrolling, be obscured, etc. This is ambiguous and therefore unsafe to rely on.
How is it that a sequence of bits representing pixels is any different than a sequence of bits representing letters and words?
Pixels are not in a sequence on the screen -- they should be processed as a single image of the screen, and sequence is formed by a changing image. It's not even possible to get ALL changes that happened over some time, GUIs are not designed for that, and should be readable for a human even if multiple changes happened between screen refreshes. Obviously, there is always some guesswork about what really got obscured, users often find that they have to "wiggle" windows around just to find if their scrollbars are really where it seems, or if there is a hidden window under them.
On the other hand, text is always a continuous stream, and it does not matter at which precise moment some piece of it is being sent -- the semantics don't change with any buffering as long as everything that is sent is eventually received.
I'm not sure how much programming you've done
More than you likely ever will.
, but you need to do more before you can authoratively say that something can't be done, especially when I can come along and say that I have done it.
Game interfaces are borderline cases that do not correspond to any traditional GUI guidelines, therefore solutions for them do not matter.
Writing a program to understand a given GUI would be fairly trivial with decent OCR and graphics libraries.
False.
Writing a program to understand a themable GUI with access to the themes and knowledge of what theme is being used would be harder, but not impossible.
False
Writing a program to understand any GUI understandable by humans would be... well impossible if you're talking about skinned MP3 players and stuff, but even a marginally consistent GUI wouldn't be impossible, just VERY difficult.
False. All those things are impossible. After a fair amount of technology development it may become possible to do them with low probability of errors, but it will be always impossible to avoid errors completely because GUIs are inherently ambiguous.
This is tolerable with humans at their terminals because ambiguity of GUIs is often outweighed by lower speed and higher probability of human errors without GUIs, however the problem remains with humans, too.
Computers do not have a preference for text. We could write a command shell that only accepted numbers, or only accepted certain tones played into a microphone, or frequencies of light, or weights, or densities, or whatever we could imagine.
All those things ARE different kinds of text, defined for the computer as sequences of bytesthat directly correspond to the data being sent across the interface.
If computers have a preference for anything, it's opened and closed circuts, and that's only if you say that computers must be a digital device.
Computers have preference for circuits, opened and closed IN SEQUENCE as means of their i/o. What is the same as sequentially written text.
Actually with a formalized legal system it would be extremely hard to convict ANYONE fro murder -- the amount and certainty of evidence will have to be far beyond what was present in OJ case. Howewer I am not talking about court proceedings that involve more hand-waving and psychological presure than anything that has to do with laws. I am talking about laws themselves and their interpretations that apply to life long before someone steps into court, plain questions about whatis legal and what is not.
Ex: anything that has to do with fair use of copyrighted material.
There's quite an oppurtunity cost for the computer illiterate to learn to program... let alone program well. It makes more sense for the computer illiterate individual to trade resources for computer software/services.
There is an even higher opportunity cost for being born, yet all people that you have ever seen, were born at some point.
People _do_ adapt to a toaster. They place bread in it, press on the lever and wait for toast to be ready.
A person that is not adapted to the toaster would demand that a toaster will emulate a pastry chef, take orders in spoken English, and give out ready toasts without burdening the user with bringing and slicing bread, placing it into the correct position, pressing levers or listening to a loud "Twang!!!" when the toast is ready.
Same about computers. People who advocate making computers adapt to humans want a computer to behave like a secretary, like a VCR, like a mathematician, physicist, engineer -- whatever but behaving like a computer. What is fine in some cases and completely wrong in others.
Computers have no preference for text. They have no preference for graphics. If they could be said to have any sort of preference at all, it would be binary. And that would still be a misleading statement.
Sure they do. Try to write a program that can operate on another computer, reading the bitmap ot its screen in GUI and sending keyboard/mouse events. Can't be done, there will be always ambiguous and hard to recognize situations (like, oddly overlapping windows and shapes, unfamiliar fonts, etc.), unless there is some severe limitation imposed on the GUI.
On the other hand, countless scripts talk to the terminals or other programs through text streams.
GUI is something that a computer program can not understand -- humans and only humans have the set of metaphors that allow them to see objects in what GUI draws on the screen. This set of metaphors is completely foreign to the computer itself, and its idea of objects is merely represented in the GUI picture.
They are only different because it is profitable for some companies to artificially limit the number of people who can do their jobs. Really there are only:
1. Data entry.
2. Secretary-like job of writing pretty letters.
3. Data processing.
4. Writing applications.
5. Writing tools for writing applications.
6. Writing infrastructure.
7. Writing tools for writing infrastructure.
First two can go feed some sharks, and I won't notice. Last four are busy enough writing code already. The third one is actually what that the question is about.
The idea that "data processing" is an uncreative job, and all such a person should do is taking pretty letters written by secretaries, demand from application programmers to write something that process them and outputs more pretty letters, and then using those applications like a good peon until there is again a reason to bitch at the application programmer, is widespread but ridiculous and insulting.
People who understand what data they deal with, should be at least able to express their procedures in a way that doesn't require a trip to the application programmer every time they need to summarize another column, or attach another database. And when those people are incapable of doing that because they were taught an extremely limited idea of how they are supposed to use a computer, the efficiency of their work drops to a ridiculous level, reducing them to the full time "data entry clerks".
Um, no.
Subsistence farming was extremely rare through the history, usually "peasant" was not just producing the food for his own family but had various kinds of obligation to "lord", "owner", "landowner", "government" that taxed him, controlled his behavior, and often property, and performed various services for him that the peasant was incapable of, lacking time, skills or education.
In other words, someone in a subservient position that requires him to perform large amount of agricultural work but prevents him from doing anything more complex.
GUI is useful. The point is, if ALL you know is GUI, you can't do anything that others didn't do for you, and you need to know more than that to do something less trivial. Sure, Photoshop (or Gimp) GUI works fine, but each and every piece of those GUI was made for a purpose that was already known at the time of its writing, and there is more to image editing than what is written and neatly packaged in those programs and plugins to them. It's a tool, and there is nothing wrong with tools, however inability to modify one's tools is always a limitation on the person's creativity, and at some point the user is faced with a situation where the tool is inadequate. This is where ignorant user has to do thousands of drags and drops, or has to train his aim to hit a particular pixel on the screen with a mouse, and knowledgeable one types few tens of characters.
There always will be something that "peasants" won't care to learn, and among that will be "literacy", something that truly differentiates them from more advanced parts of the society. Modern "peasants" may not be defined as people who do farming, and are placed under the rule of their lords, but there certainly there is a part of the society that survives while getting inadequate education, and having ridiculously low amount of decisions that they can make for themselves.
If something will improve this situation, it will be the elimination of the things that cause large number of people to become "peasants", however I am not holding my breath for that. In particular, modern American society's economy, politics and culture are entirely based on the presence of a large uneducated, faceless, easy to manipulate mass right above the bottom of the society and below the much smaller number of well-educated people.
Eventually when more complex knowledge will become a thing that separates peasants and lords, some of the current "lords" may find their heirs being thrown to the "peasants" level, and some "peasants" may find themselves "accidentally" acquiring literacy as a side effect of job training or as a result of their curiosity, but it's unlikely that the structure will get fundamentally changed because of that. It can only be changed if there will be a force that purposefully acts to eliminate the "illiteracy", and therefore the existence of "peasants" and their servitude. Too bad, American society has no such force.
There is one area of everyday human activity where complex logical structures are always expressed using natural language without any aid from more formalized languages and systems of expression.
It's called legal system, and it's a mess so huge, no one, human or machine, can predict its behavior, and an army of lawyers, judges and lawmakers of all kinds spend untold amounts of time trying to implement it in their various ways. If computer programming will ever turn into something that resembles this, determining which color should be a pixel (15, 351) ten seconds into a Quake game will be a process not unlike the Kobe Bryant case.
If everything is a stream of bytes, any kind of interface can be created with it, based on its purpose. Simple interfaces can hide complex structure by just exchanging blocks of data blindly, complex ones can use parsers (a thing that Windows programmers never learn) and implement any format easily. And this clearly separates the programs -- they never ever touch each other's address space and can have any, simple or complex, design of their objects. Also whatever can be done with pipes, can be done with sockets, thus allowing remote and distributed processing without any change to the software doing it.
Dotnet's "command line", judging by the descriptions that Microsoft made, allows things to just sit in one happy clusterfuck and mess with objects while they are freely floating inside the shell application, something that is in no way different from writing "scripts" in a large spreadsheet. This is yet another example how Microsoft invents 65537th iteration of the mix of their original DLL and DDE, two ideas that they still can't get right after more than a decade of development.
And not a single word about source and messages' leak that demonstrated how bad the system really is. Instead we see claims that Diebold has "empirical data" that voters love the new machines just like Happy Meals at Mcdonalds.
So basically Microsoft, in their fine tradition of creating an confusion in terms, made a "shell" that can't run programs but instead runs "managed code" (aka shared libraries in dotnet environment) and maintains all objects inside itself, but has a syntax that is confusingly similar to Unix shell pipes. The concept is obviously flawed -- Unix shell uses Unix-specific unified file descriptor model, that Windows lacks, so while in Unix shell is often used to process easily-parseable text using pipelines of programs (some remote) and any overcomplication of data passed between programs is frowned upon, in Windows the most convoluted ever format is being used to exchange things between "programs" that are sitting in the belly of that "shell". They could just as well make the whole interface an UML editor.
If anything, it can prevent some people who learned that shit from switching to Unix -- they will until the end of their miserable lives associate pipelines with shitloads of DLLs instead of streams of text.
This is just barking up the wrong tree. Of course, there is a decline, but mesauring the accomplishment by the amount of standing out individuals is a wrong way to measure or find a decline. In general the amount of "celebrity cult" of scientists, politicians and writers through history decreases, as they become more numerous, and their accomplishments require more co-operation. In politics it can be seen when one compares absolute monarchies of the past when almost all political acts are attributed to a single person, to modern systems with large governments. However the same can be said about science, except that in a modern world when a group of politician starts a war, most of their names usually get known and mentioned, but when a group of scientists develops a powerful weapon, most of names are not mentioned, and certainly nothing is known about people who re-done the same work elsewhere.
Also different cultures have differrent traditions of crediting individuals to begin with, not to mention that large amount of material is published in the languages other than English/German/French/Spanish/Italian, and therefore is not being referenced by European sources, so the analysis of the "accomplishment" distribution through the world is even more skewed than the distribution of them through the time.
So the actual amount of progress beind made can't be measured by just individuals that stand out enough to get credited, at some point more accomplishment leads to less visible "leaders". However the decline in overall progress started in the second half of the 20th century is definitely intuitively visible. In my opinion, the problem is unrelated to personal achievements, and is based on the skewed and distorted structure of the global economy that formed at the same time. Basically, industrial revolution was based on the idea that if mechanisms won't increase people's productivity, nothing else will. When it became easier to "let someone else do this", the incentive to develop similar kinds of advances in science and engineering disappeared. As an example I can see it as great dishonor to humankind that my Levi's 501 jeans that I am wearing (and went through at least 10 pairs over my life so far) are not being produced on a humanless automated line, employing CNC cutting and sweing machines exclusively. It would need a pretty small amount of innovation in engineering, and maybe some work in science (machine vision that would be able to handle fabric texture properly, see threads and seams at odd angles, etc.). But it's not being done because sweatshops are cheaper, even if they are a horrendous waste of human effort otherwise.
The progress in technology became strongly skewed toward entertainment-related purposes because they provide higher return on investment compared to similar effort made "sweatshop-style", however I doubt that there is a natural process that can extend those advances into other areas until the global economy stabilizes to make domestic industry a necessity for all countries again. That would basically mean a return of industriak revolution, though likely under other name, and with a lot of pundits claiming that it's something "even better than our postindustrial world". In my opinion, "postindustrial" societies are merely declined elite, fed by sweatshops, not unlike arsitocracy/peasants-based feudal societies that preceded the industrial revolution in the first place.
Thought recognition technology has actually gone a lot further than what you imply (I first saw a report on this sort of thing 2-3 years back).
This is barely any recognition at all -- it uses signals from few neurons, maps them to movements parameters, and allows visual feedback. Feedback also simplifies the procedure by allowing the brain to "re-learn" how it is supposed to control the movement, as opposed to device learning how brain does that already.
It can be called "thought" only because it's not a signal collected from nerves going to the muscles, but its predecessor, however it contains the same amount and kind of information. No symbolic or abstract information is passed, so it amounts to giving the user an equivalent of joystick or mouse, that is harder to control than an actual joystick or mouse, a thing that never was a problem for a wearable computer in the first place. Again, beyond that level the problem is fundamental, and would require multiple levels of recognition, each involving reverse-engineering a mechanism that we know almost nothing about.
Even if you regard the above as "most basic functions", that functionality is more than good enough for what I'm talking about. Of course the current technology is quite invasive and drastic. So I don't see that being very popular anytime soon.
Also it's still impossible to get this signal without damaging large number of neurons in the process of installing the sensors, and the lifetime of electrode is limited.
Now if they get to the stage of giving people SVGA/s-video inputs, things could get even more interesting.
Actually there were some successful tests of prosthetic vision devices. Just as invasive, produces a poor resemblance of an image, and involves re-learning of the image processing in the brain to whatever mapping that the device produces, as opposed to imitating the mapping originally used to process signals from the eyes (that, again, is very little known about beyond the basic encoding principles that the device imitates).
Two cameras (one looking forward, another looking at the pupil), and one in-glasses projection display per eye will provide the same functionality better, and a chord keyboard, or twiddler-like single-hand keyboard already can provide good enough symbolic input, at least until some fundamental discovery of the processing/reverse engineering procedures for thought mechanisms will be made.
But most of the stuff I mentioned is doable now.
It is not. If handwritten text recognition from a photograph can't be done, an part of your device's purpose is lost -- you can only manually convert those photographs to text, and unless you are in a hurry when taking those, and have plenty of time to convert them to text, you gain nothing compared to just taking notes. It will be an equivalent of using an audio record of a lecture to make notes later -- often a good compromise but certainly not a replacemnt for instant recording. And since there are fundamental problem with this kind of "handwriting recognition", the solution isn't going to be available any time soon.
The rest can be added later - e.g. thought pattern recognition (already done with rats, apes and monkeys),
Those only could "recognize" the most basic functions, at the expense of very intrusive sensors and large measurement equipment. Again, there are fundamental problems with higher levels of functionality, and no noticeable progress on those was made in at least 20 years, placing this work in the same category as fusion energy. So not any time soon, either.
face recognition (this doesn't have to be 100% accurate - coz you can quickly verify it's your friend/enemy/customer/boss) and so on.
Face recognition is so far from those tasks, ans can be reduced to such a simple statistical manipulations on geometry, it's primitive compared to those tasks. The fact that it's still unreliable only demonstrated how far we are from the solution for the rest of those problems.
There's a reason for that, and it's tied into something that a lot of people here on slashdot and similar tech-based sites don't understand: most people don't want to have to learn something new just to get their job done.
If they don't have sufficient knowledge to do their work, doesn't it mean that they are incompetent? Sure, if there are alternatives that don't require such a knowledge while still letting them do their job, no (or little) harm is done, but so far most of the attempts to make an executive's work easier to perform without requiring any new knowledge, failed miserably. The nature of work that involves dealing with large amounts of text, numbers and various representations of those, absolutely demands an ability to read and write them.
Substituting text with "digital ink" that is an inferior substitute of text (can't be processed, hard to read), and numbers with the same "digital ink" that is a completely inadequate replacement for numbers (recognition errors can't be easily found and fixed), leaves people who use those just as incompetent as if they used pen and paper exclusively.
Executives, as you point out, don't need to type because they "have someone to do that for them".
If this is true, and an executive is capable of doing his work with all text input being done by secretary, he gains nothing from replacing a secretary with a pad that he writes squiggles on. Usually the real problem is that the job demands closer interaction with the data, and a secretary can't do all that efficiently. But then the solution offered should be more efficient than a secretary, not less efficient like the tablet design is -- then it's better to just keep a secretary.
A lecture hall of 250 students waiting a whole minute for acrobat reader to load is not fast enough.
The loading time of Acrobat Reader has little to do with the CPU speed, and A LOT with the use of a slow notebook hard drive (in this case even designed to work when being rapidly tilted), and RAM+swap being filled with crap.
There is a big difference between:
- handwriting recognition that takes pen movement as an input, and allows user to correct the mis-recognized text (this is what everything uses -- with moderate success),
- handwriting recognition that uses pen movement as an input without a feedback (what can work, but usually only after either user or computer will go through a lot of learning in a feedback mode),
- "handwriting recognition" that takes the images of written text as an input (what is something that even humans often have problems with, and for the modern recognition technology is still in a "don't even try" area).
Actually you will make a potent TARGET -- the probability of it being hit nears 1 in any imaginable battle because no one among the enemy will shoot at anything else until it will be down.
If you think about it, the only way to "abolish war" is peace through superiority. Nobody will directly confront an obviously superior opponent.
I would advise the military of all nuclear-capable countries, other than US, to once in a year nuke a city where the person, who made the most recent statement of this kind, resides, until everyone gets the idea. You know, to show how wrong they are.
The article is a great example of marketspeak run amok. Really, nothing to see other than the consequences of reading too much trade rags.