We're already at a point of labor oversupply. Anyone with a basic understanding of economics knows what happens to the price of something when supply exceeds demand. As robots replace demand, we need desperately to come up with new stuff for people to do, or we're going to see the middle class completely vanish. After all, the modern American middle class was built on factory jobs that no longer exist. Technology has displaced old and created news jobs before, but this presents a challenge completely unprecedented in which unskilled labor becomes completely useless.
If that happens, we've got two choices: (1) Let those outside of the top 10% struggle because they aren't useful anymore. (2) Redistribute wealth to provide for basic necessities and let work become more voluntary.
Option #3: Decentralize high-tech production the same way low-tech work was done some 300 years ago. Every town had its own carpenter, blacksmith, shoemaker, tanner, tailor etc. In the near future, every town will need a small workshop with basic CNC machines and 3D printers capable of producing at least a 1990s-level computer, fridge and car engine. If we manage to get rid of copyright and patent monopolies, this is the path of least resistance.
When was the last metal you saw that was less dense than water? Surely, with enough time and effort we can separate metal powders, we did originally separate them from rocks after all. I don't think it'd be worth it though.
There are other liquids much denser than water, you know? If you won't insist on a particular temperature range, anything can be liquid.
Instead of changing materials once, abruptly, creating a brittle failure plane, couldn't you change materials a whole bunch of times little-by-little (changing the alloy mixture) so that you get a gradient of material properties?
That depends a lot on the materials in question. If the materials don't mix well, the best you'll get by mixing the powders in a progressive fashion would be very close to one metal with lots of impurities. It would be a little stronger than straight boundary but any impurity is still a perfect starting point for fractures.
I work with a lot of different rapid prototyping processes - including DLMS - on a weekly basis. In their current form you cannot change materials mid-part on a DLMS machine. Even if the machine itself could handle it (i.e., had multiple material-handling streams), you would have a tough time getting the dissimilar metals to properly fuse. As a welder how good the results of welding aluminum and steel are, or titanium and steel. Mixing these metals tends to result in brittle intermetallic phases, which are perfect places for fracture. Even worse is that, because the machine works layer-by-layer, the boundary between the two metals would be planar, making fracture all that more likely.
Actually, if you could deliver multiple types of powder into the print area with reasonable precision (normal office laser printers can print on paper with 1200dpi precision, though using electrostatically charged roller like in a normal laser printer may not be applicable for 3D printing), you could solve the problem of (not) fusing the metals together by printing a complicated interlocking 3D pattern around the boundary. While the metals may not fuse anyway, it won't matter because the brittle boundary will not be strained by the load at all. The interlocking structure of two pure solid metals will bear all the mechanical strain instead.
What is more, because the process is akin to SLS, what you end up with is a box - the build envelope - completely filled with powder, inside of which is the part you've made. You can ordinarily reclaim most of the powder and use it again. If you were to change materials mid-stream, you would have to junk most if not all of the leftover powder, because you wouldn't have a good way to separate the powders.
There are lots of ways to separate different materials. Ferromagnetic vs. non-ferromagnetic - use a magnet. Very different density - put them in a liquid where one will float on the surface and the other drop to the bottom. And so on.
When making metal parts the the metallurgical properties are more important than the shape. The direction of the molecules in the metal make the difference between something that's hard and brittle vs. something that's soft and elastic.
You can harden and temper the printed metal part just like any other.
Coding sucks today because our programming languages and our compilers require more machine understanding than problem solving ability.
And eventually that will change.
But today, the mastering the tools and intricacies is FAR MORE IMPORTANT than actually being able to describe the problem.
And like I said, eventually intuitive problem solving applications will emerge and trump programming languages.
I shouldn't have to know the scope of a variable, whether I want to allocate some memory, how many bytes and whether or not I want it static to solve a problem. But today, our computers aren't actually really that fast either. They just seem fast compared to the snail speed ones of the past. But someday....
So... you basically want to perfectly speak a foreign language without bothering to learn its grammar and vocabulary. Because that's what programming languagesARE: man-made foreign languages focused on expressing ideas with extreme precision.
Every single brilliant attempt to free programmers from restraints of resource management and abstract programming language constructs ended up as complete and utter failure where doing anything more complicated than adding a few numbers together gives you massive headache. And believe me I've seen lots of such attempts.
What a load of bullshit. Understand this, whatever means you have YOU CANNOT FIGHT the government on technical solutions alone.
I never said I could. The technical solutions are the bare minimum necessary to keep us in the fight at all.
That's why laws are important and why the Constitution is important. Once the government starts disregarding its own laws, or worse coming up with arbitrary laws so to circumvent the protections afforded to you by the Constitution it's game over. You get it dumbass ? It's game over. Guns won't protect you, encryption won't protect you. Nothing will except good old fashioned written words on some old document 200 years old.
What's going to protect those 225 year old words from the government, then? That's right, guns and new words.
That line of thinking is worse than complacency, it's acceptance of corruption.
It's not acceptance of corruption, it's damage management. First, make sure in advance that possible attacks won't cause any serious damage. Then, after the attack actually happens and fails, fight back as hard as you can. If you allow the attack to succeed, you may lose the ability to fight back at all.
Enjoy Obummer's America you fucking retarded liberals.
All praise to the Glorious Leader!
More like enjoy the voting-against-the-other-guy America. Red vs. Blue politics completely devoid of any ideas or alternatives. When the worst case scenario for the ruling party is to play minority for the next one or two terms no matter how badly they screw up, don't expect your rights and freedoms to last long.
Perhaps. But that doesn't change the fact that we all have to turn up our paranoia level way up just to keep our ability to fight for our freedoms. If we don't, we'll never get back the freedoms we've already lost in the past few decades and we'll lose even those that we still have today.
And tell me, you think most people out there have the materials science knowledge, the engineering skills and the patience to diddle around with this stuff?
The main advantage of 3D printing is that it's significantly simpler and easier to use than CNC machines.
What will change politically and economically? It's nothing more than a hobby, and how many people do you know that have a fully-equipped CNC shop at home, and what did it change for them politically or economically?
So were personal computers in 1980s. As expertise requirements go down, adoption by the general public goes up.
The DSL router itself could be compromised as well. I'd start by booting up a Linux live CD, disconnecting everything else from the network and changing the external IP address again. Then I'd wait to see if they find you again. If they don't, start plugging everything back one device at a time, again checking if they find you after plugging the last device in.
Well, in a free-market situation, every career path is a low-paying one, since money in your pocket is an inefficiency (from the market's point of view) which competition will remove. Thus getting an education is always a poor life choice there, having an expected utility that tends towards zero. And so does every other choice; that's what efficient market means.
This is not true if human labour is the main limiting factor of production. Limiting factors are always expensive while abundant resources are always cheap. The problem is that due to technological progress, human labour has become a cheap abundant resource while the biggest limiting factors are some natural resources and, surprisingly, money itself. That's why the financial markets have grown so ridiculously huge compared to the tangible economy.
And, none of this exclusively applies to the unskilled (though, obviously, the marginal cost of production of a college-educated worker is a lot higher, so such workers must earn more in absolute terms in order for their price to be equal to their marginal cost of production).
When fresh graduates start with huge debt, the end result is precisely the opposite. The debt forces them to accept even jobs where they'll be grossly underpaid. After all, why would the employers care about debts of their employees?
Thanks. This research doesn't translate directly for 3D printing (3D printers are usually designed for solid filament) but you could build a special 3D printer with additional inkjet head to print the circuits. I guess that will be the best way to achieve the necessary precision for microchips.
Can you get a 3D printer to even just print out a standard 7400 NAND gate? That seems like quite the leap of technology to get even that far, and would seem to be at least on the threshold of possibilities. So far, I've never seen even something like that.
Not yet because I don't have the necessary materials. But I expect that somebody will discover usable materials for 3D printing of semiconductors within the next 5 years. After that, going from individual transistors to simple logic gates is a matter of days (specifically, TTL NAND gate is built from 2 transistors, one of which has multiple emitters).
There even would be a pretty active hobbyist market if you could get a few thumb-sized chips printed out in that manner.
No, it's not. "Conventional" 3D printers can't make transistors at all, and it has nothing to do with the minimum size of the features they can print. It's about what transistors are made of, the processes used to make them, requirements for clean facilities to avoid materials contamination (it's not just about defects), and so on.
As is usual for 3D printing stories, starry eyed enthusiasts with absolutely no fucking clue about how physical things are made are out in force claiming that everything!!1!1!11!!!! is going to be 3D printed in the home soon1!!1!!1!. I regret to inform you that this is the role you are currently playing. You can imagine a thing all you like, but your imagination is not a reliable guide to reality.
Conventional transistor / IC manufacturing processes use high-tech versions of lithography, i.e. they're a form of 2D printing. Three dimensional structures (of layered thin films) are built up with successive 2D printing steps, and many of the steps involve selective subtraction (e.g. print a mask, perform an additive step, then strip the mask). While you could in principle just build an additive "print head" akin to the type of 3D printer you're so unrealistically enthusiastic about, nobody would actually use this for production because it's so much less economical than lithography. (Thanks to what is essentially parallelism -- lithographic printing forms an image across the entire plane in one step, instead of requiring a head to slowly raster-scan everything. Also, optically printing the entire plane in one step gives better control over dimensions and tolerances than a raster-scan device can.)
As long as you can put two differently doped pieces of the same semiconductor material together so that a P-N junction forms, you can build a working transistor. It doesn't matter how you do it. Controlling the properties of finished transistors is a different question and completely besides my point.
3D printer is much easier to operate than a milling machine. You don't have to sharpen any tools, you don't have to switch tools and you don't have to calibrate the starting position over the piece of raw material every time you change the program. With a 3D printer, you just put the raw material into the printer's feeder, open the design file you want and click "print".
There is a reason IC plants are so expensive; you can't print a chip without a lot complex machinery, a specific environment, etc. So, even if somebody comes up with a printer that can print a laptop it will have a large up front cost, require maintenance, and not be cheaper than paying a company that specializes in making laptops for many decades to come.
The reason is the size of individual transistors. Making a chip with 14nm transistors is damn expensive and probably way beyond what conventional 3D printers could ever achieve. Making a millimeter-sized transistors is cheap and within capabilities of current 3D printers. I can imagine that 3D printers could print something like the good old Zilog Z80 within 10 years.
No, the reason why this isn't being fixed is that a significant number of voters vote for politicians that run on a platform that includes deregulation and freeing the market of even modest restraints on bad behavior. And are quite vocal in shouting down anybody that suggests even modest reforms as being elitists and pushing for a totalitarian, nanny state.
Those people won't know what "totalitarian" means until the police state they have been voting for all along under the pretense of war against terrorism kicks down their front door and drags them away with a black cloth bag over their heads never to be seen again.
Slashdot Mirror
We're already at a point of labor oversupply. Anyone with a basic understanding of economics knows what happens to the price of something when supply exceeds demand. As robots replace demand, we need desperately to come up with new stuff for people to do, or we're going to see the middle class completely vanish. After all, the modern American middle class was built on factory jobs that no longer exist. Technology has displaced old and created news jobs before, but this presents a challenge completely unprecedented in which unskilled labor becomes completely useless.
If that happens, we've got two choices: (1) Let those outside of the top 10% struggle because they aren't useful anymore. (2) Redistribute wealth to provide for basic necessities and let work become more voluntary.
Option #3: Decentralize high-tech production the same way low-tech work was done some 300 years ago. Every town had its own carpenter, blacksmith, shoemaker, tanner, tailor etc. In the near future, every town will need a small workshop with basic CNC machines and 3D printers capable of producing at least a 1990s-level computer, fridge and car engine. If we manage to get rid of copyright and patent monopolies, this is the path of least resistance.
When was the last metal you saw that was less dense than water? Surely, with enough time and effort we can separate metal powders, we did originally separate them from rocks after all. I don't think it'd be worth it though.
There are other liquids much denser than water, you know? If you won't insist on a particular temperature range, anything can be liquid.
On the other hand, you can print thousands of unique parts before you prepare one new forging die.
Instead of changing materials once, abruptly, creating a brittle failure plane, couldn't you change materials a whole bunch of times little-by-little (changing the alloy mixture) so that you get a gradient of material properties?
That depends a lot on the materials in question. If the materials don't mix well, the best you'll get by mixing the powders in a progressive fashion would be very close to one metal with lots of impurities. It would be a little stronger than straight boundary but any impurity is still a perfect starting point for fractures.
I work with a lot of different rapid prototyping processes - including DLMS - on a weekly basis. In their current form you cannot change materials mid-part on a DLMS machine. Even if the machine itself could handle it (i.e., had multiple material-handling streams), you would have a tough time getting the dissimilar metals to properly fuse. As a welder how good the results of welding aluminum and steel are, or titanium and steel. Mixing these metals tends to result in brittle intermetallic phases, which are perfect places for fracture. Even worse is that, because the machine works layer-by-layer, the boundary between the two metals would be planar, making fracture all that more likely.
Actually, if you could deliver multiple types of powder into the print area with reasonable precision (normal office laser printers can print on paper with 1200dpi precision, though using electrostatically charged roller like in a normal laser printer may not be applicable for 3D printing), you could solve the problem of (not) fusing the metals together by printing a complicated interlocking 3D pattern around the boundary. While the metals may not fuse anyway, it won't matter because the brittle boundary will not be strained by the load at all. The interlocking structure of two pure solid metals will bear all the mechanical strain instead.
What is more, because the process is akin to SLS, what you end up with is a box - the build envelope - completely filled with powder, inside of which is the part you've made. You can ordinarily reclaim most of the powder and use it again. If you were to change materials mid-stream, you would have to junk most if not all of the leftover powder, because you wouldn't have a good way to separate the powders.
There are lots of ways to separate different materials. Ferromagnetic vs. non-ferromagnetic - use a magnet. Very different density - put them in a liquid where one will float on the surface and the other drop to the bottom. And so on.
But weaker parts.
When making metal parts the the metallurgical properties are more important than the shape. The direction of the molecules in the metal make the difference between something that's hard and brittle vs. something that's soft and elastic.
You can harden and temper the printed metal part just like any other.
Coding sucks today because our programming languages and our compilers require more machine understanding than problem solving ability.
And eventually that will change.
But today, the mastering the tools and intricacies is FAR MORE IMPORTANT than actually being able to describe the problem.
And like I said, eventually intuitive problem solving applications will emerge and trump programming languages.
I shouldn't have to know the scope of a variable, whether I want to allocate some memory, how many bytes and whether or not I want it static to solve a problem. But today, our computers aren't actually really that fast either. They just seem fast compared to the snail speed ones of the past. But someday ....
So... you basically want to perfectly speak a foreign language without bothering to learn its grammar and vocabulary. Because that's what programming languages ARE: man-made foreign languages focused on expressing ideas with extreme precision.
Every single brilliant attempt to free programmers from restraints of resource management and abstract programming language constructs ended up as complete and utter failure where doing anything more complicated than adding a few numbers together gives you massive headache. And believe me I've seen lots of such attempts.
What a load of bullshit. Understand this, whatever means you have YOU CANNOT FIGHT the government on technical solutions alone.
I never said I could. The technical solutions are the bare minimum necessary to keep us in the fight at all.
That's why laws are important and why the Constitution is important. Once the government starts disregarding its own laws, or worse coming up with arbitrary laws so to circumvent the protections afforded to you by the Constitution it's game over. You get it dumbass ? It's game over. Guns won't protect you, encryption won't protect you. Nothing will except good old fashioned written words on some old document 200 years old.
What's going to protect those 225 year old words from the government, then? That's right, guns and new words.
That line of thinking is worse than complacency, it's acceptance of corruption.
It's not acceptance of corruption, it's damage management. First, make sure in advance that possible attacks won't cause any serious damage. Then, after the attack actually happens and fails, fight back as hard as you can. If you allow the attack to succeed, you may lose the ability to fight back at all.
Fascism begins when the efficiency of the Government becomes more important than the Rights of the People.
Exactly. The government being slow and cumbersome is not a bug of democracy, it's a damn important feature.
Enjoy Obummer's America you fucking retarded liberals.
All praise to the Glorious Leader!
More like enjoy the voting-against-the-other-guy America. Red vs. Blue politics completely devoid of any ideas or alternatives. When the worst case scenario for the ruling party is to play minority for the next one or two terms no matter how badly they screw up, don't expect your rights and freedoms to last long.
I don't suppose that you noted that the files in question were (believe it or not) on dead plants. Every tried encrypting dead plants?
There's a handy tool for that invented way back in the stone age. It's called "fire". Works like a charm. It's a one-way process, though.
You are blaming the victim here.
Perhaps. But that doesn't change the fact that we all have to turn up our paranoia level way up just to keep our ability to fight for our freedoms. If we don't, we'll never get back the freedoms we've already lost in the past few decades and we'll lose even those that we still have today.
And tell me, you think most people out there have the materials science knowledge, the engineering skills and the patience to diddle around with this stuff?
The main advantage of 3D printing is that it's significantly simpler and easier to use than CNC machines.
What will change politically and economically? It's nothing more than a hobby, and how many people do you know that have a fully-equipped CNC shop at home, and what did it change for them politically or economically?
So were personal computers in 1980s. As expertise requirements go down, adoption by the general public goes up.
The DSL router itself could be compromised as well. I'd start by booting up a Linux live CD, disconnecting everything else from the network and changing the external IP address again. Then I'd wait to see if they find you again. If they don't, start plugging everything back one device at a time, again checking if they find you after plugging the last device in.
Well, in a free-market situation, every career path is a low-paying one, since money in your pocket is an inefficiency (from the market's point of view) which competition will remove. Thus getting an education is always a poor life choice there, having an expected utility that tends towards zero. And so does every other choice; that's what efficient market means.
This is not true if human labour is the main limiting factor of production. Limiting factors are always expensive while abundant resources are always cheap. The problem is that due to technological progress, human labour has become a cheap abundant resource while the biggest limiting factors are some natural resources and, surprisingly, money itself. That's why the financial markets have grown so ridiculously huge compared to the tangible economy.
And, none of this exclusively applies to the unskilled (though, obviously, the marginal cost of production of a college-educated worker is a lot higher, so such workers must earn more in absolute terms in order for their price to be equal to their marginal cost of production).
When fresh graduates start with huge debt, the end result is precisely the opposite. The debt forces them to accept even jobs where they'll be grossly underpaid. After all, why would the employers care about debts of their employees?
The market is a legal fiction. Aka a customary practice of a certain group of human beings.
Makes no difference.
Thanks. This research doesn't translate directly for 3D printing (3D printers are usually designed for solid filament) but you could build a special 3D printer with additional inkjet head to print the circuits. I guess that will be the best way to achieve the necessary precision for microchips.
Can you get a 3D printer to even just print out a standard 7400 NAND gate? That seems like quite the leap of technology to get even that far, and would seem to be at least on the threshold of possibilities. So far, I've never seen even something like that.
Not yet because I don't have the necessary materials. But I expect that somebody will discover usable materials for 3D printing of semiconductors within the next 5 years. After that, going from individual transistors to simple logic gates is a matter of days (specifically, TTL NAND gate is built from 2 transistors, one of which has multiple emitters).
There even would be a pretty active hobbyist market if you could get a few thumb-sized chips printed out in that manner.
My point exactly.
No, it's not. "Conventional" 3D printers can't make transistors at all, and it has nothing to do with the minimum size of the features they can print. It's about what transistors are made of, the processes used to make them, requirements for clean facilities to avoid materials contamination (it's not just about defects), and so on.
As is usual for 3D printing stories, starry eyed enthusiasts with absolutely no fucking clue about how physical things are made are out in force claiming that everything!!1!1!11!!!! is going to be 3D printed in the home soon1!!1!!1!. I regret to inform you that this is the role you are currently playing. You can imagine a thing all you like, but your imagination is not a reliable guide to reality.
Conventional transistor / IC manufacturing processes use high-tech versions of lithography, i.e. they're a form of 2D printing. Three dimensional structures (of layered thin films) are built up with successive 2D printing steps, and many of the steps involve selective subtraction (e.g. print a mask, perform an additive step, then strip the mask). While you could in principle just build an additive "print head" akin to the type of 3D printer you're so unrealistically enthusiastic about, nobody would actually use this for production because it's so much less economical than lithography. (Thanks to what is essentially parallelism -- lithographic printing forms an image across the entire plane in one step, instead of requiring a head to slowly raster-scan everything. Also, optically printing the entire plane in one step gives better control over dimensions and tolerances than a raster-scan device can.)
As long as you can put two differently doped pieces of the same semiconductor material together so that a P-N junction forms, you can build a working transistor. It doesn't matter how you do it. Controlling the properties of finished transistors is a different question and completely besides my point.
Can a 3D printer produce clean drinking water,
Yes, if you have the materials to print a water filter or solar water still.
reliable electricity,
Yes, if you have the materials to print a solar panel (I estimate it will take about 5 years until somebody successfully prints a primitive one).
law and order and competent, honest government?
Maybe.
3D printer is much easier to operate than a milling machine. You don't have to sharpen any tools, you don't have to switch tools and you don't have to calibrate the starting position over the piece of raw material every time you change the program. With a 3D printer, you just put the raw material into the printer's feeder, open the design file you want and click "print".
There is a reason IC plants are so expensive; you can't print a chip without a lot complex machinery, a specific environment, etc. So, even if somebody comes up with a printer that can print a laptop it will have a large up front cost, require maintenance, and not be cheaper than paying a company that specializes in making laptops for many decades to come.
The reason is the size of individual transistors. Making a chip with 14nm transistors is damn expensive and probably way beyond what conventional 3D printers could ever achieve. Making a millimeter-sized transistors is cheap and within capabilities of current 3D printers. I can imagine that 3D printers could print something like the good old Zilog Z80 within 10 years.
No, the reason why this isn't being fixed is that a significant number of voters vote for politicians that run on a platform that includes deregulation and freeing the market of even modest restraints on bad behavior. And are quite vocal in shouting down anybody that suggests even modest reforms as being elitists and pushing for a totalitarian, nanny state.
Those people won't know what "totalitarian" means until the police state they have been voting for all along under the pretense of war against terrorism kicks down their front door and drags them away with a black cloth bag over their heads never to be seen again.