"Actual" or "General" Intelligence is irrelevant. As is "understanding", "comprehension", "self awareness" and any other hurdles you put in the definition of "True AI".
What is relevant, is that we are making machines capable of displaying sophisticated "identify a square" levels of pseudo-"intellectual" achievement. We can now offload new and ever-growing classes of task that used to require human intelligence to a machine, where it can be accelerated and parallelized to a degree completely financially infeasible when relying on humans. The refinement of the mechanical calculator doesn't begin to compare to what we have unlocked.
The robots responsible for doing the majority of trading on the stock market anymore do not understand what they're doing - but that doesn't actually matter. They can still process a far vaster sea of information than a human mind could begin to absorb, and find patterns to guide them faster and more reliably than a human. Usually. It's that failure mode that's one of the other big problems - because they don't actually understand what they're doing, they don't just stumble, they go off the rails at full throttle. And especially in a world where personal profits have long been divorced from personal responsibility, that sort of "win for me, right up until you lose for everybody else" bargain is *incredibly* tempting.
>acquaintance with real facts will, in the end, be better for all parties.
Hogwash. The public's knowledge of the ease and methods of adulterating milk dealt a heavy blow to the profits of unscrupulous milkmen, it was certainly not better for them.
And unlike milk, the potential for abuse of AI is disproportionately focused on the most wealthy and powerful individuals who can afford the hardware necessary to leverage it most effectively. How dare you suggest that such upstanding individuals, the very backbone of our capitalist government, should be subjected to the sort of public scrutiny that common knowledge of the ease with which they might abuse their position would no doubt engender. How dare you sir!
It's not the intelligence of machines directly we need to fear - there doesn't seem to be any immediate threat of true autonomy / free will any time soon. It's the fact that their limited intelligence has no ethical limitations, and can be harnessed to easily enable things that would be prohibitively expensive to do any other way.
As one example: cameras on every street corner - several countries have done that already, and the results are a bit unsettling to anybody who has ever read 1984, but the actual potential for abuse is limited, because there's just too much information to sift through efficiently. Feed all those A/V feeds through an AI that can use facial/gate/etc. recognition and speech analysis to create an easily searchable database of both the real-time and historical movements and public conversations of every person in the country, with anything "suspicious" automatically red-flagged - suddenly 1984 is just an authoritarian administration away.
Or for a more violent dystopia - autonomous weapons platforms, up to and including robotic soldiers/police will allow whoever controls them to crack down on the population just as hard as they want, with none of the (limited) constraints of using human soldiers who may eventually rebel at massacring innocent civilians on a regular basis.
I agree with you about the futility of trying to limit information though - at best you can keep it out of the hands of the general public, but there's no way you're keeping it out of the hands of the wealthy/powerful individuals most likely to abuse it.
True - though it depends very much on what part of the job needs human skills. For example, if you need a human to take a package up the stairs and ring the doorbell for delivery, then it probably doesn't do you much good to automate the delivery truck and render the delivery person into an idle passenger riding between stops.
Well, nothing's going to infect it while it's offline, and if you can't trust your retrieval process, then you likely have bigger problems on your hands. It is nice that write once media protects you from corruption during the retrieval process though - at least assuming you're not using a write-capable drive to read the backups: even if you can't meaningfully update write-once media, it's still generally easy enough to intentionally corrupt it.
I have to agree whole-heartedly about SSDs though - flash stores data as charge-levels on capacitors that start losing charge the moment the data is written. It may be a lot more stable than DRAM, but you couldn't convince me to trust it for archival purposes.
It's easy enough to create an oxygen-free atmosphere here on Earth - and yet we don't work metal that way here either. Perhaps there's a good reason?
First I assume you mean *melting* the iron, rather than boiling it - trying to precisely work with a gas tends to be a laboratory exercise, not an industrial one. Moreover, gaseous iron isn't magnetic, in fact the first magnetic gas was only discovered in 2009, and it was lithium cooled to within a billionth of a degree of absolute zero. And actually, molten iron isn't magnetic either - metals pretty much all lose their magnetic properties well before they melt.
Molten iron is still paramagnetic though, so you could still manipulate it with sufficiently strong magnetic fields. No longer a simple task though, especially considering that whatever iron structure you're building *will* be magnetic, and thus react far more strongly to your magnetic "nozzle" than the molten "ink" does. Plus, if you're spraying liquid metal around it's going to be almost impossible to keep it from splashing, and thus building up on everything else in the area in a chaotic fashion, including your manufacturing equipment.
As for micrometer precision - I think you'd be greatly disappointed. Completely aside from the difficulties of creating such precision, you'd need an extremely narrow stream of metal for than kind of precision to even matter - which means your nozzle has to be very close to the target to keep the stream from solidifying long before it makes contact. Vacuum insulates you from conductive heat losses, but does nothing against radiant losses. So you're basically back to a boring 3D printer, and all the magnetic guidance stuff just introduces needless complexity.
And finally, even assuming all the other problems could be solved, there's one big honking one that probably wont be: material strength. All that hammering and crushing of iron isn't just to get it into a specific shape, it also imparts far greater strength than is present in the original, brittle, cast-iron stock. A very large percentage of the strength and durability of iron products is due to the internal molecular alignments created through working it. Even with today's high-end 3D metal printing, there's usually a final firing process designed to *almost* melt the structure so that the layers merge together at the molecular level, dramatically increasing strength. And even with that, you'd only be looking at solid-cast strength, at best. Worked strength is likely to be permanently out of reach of 3D printing.
Not at all. All it assumes is that making things out of iron would be useful. More sophisticated alloys and manufacturing techniques would undoubtedly be developed too - but we have to start somewhere, and iron is an incredibly useful material just as it is, and far easier to work with when we're just starting out.
As we expand into space, it makes sense to pick the low-hanging fruit first. Bulk construction using iron, radiation shielding using "cemented" dust and gravel, and rocket fuel from water - the technologies for all three are reasonably straightforward, and they address the most mass-intensive demands for developing infrastructure in space.
So where's the disagreement? As I said, the license code off an old PC case sticker is almost certainly going to be useless for anything else. And without a valid license code, so will that shiny new copy of Windows you just downloaded from Microsoft.
No copyright infringement or license violations involved, unless you get lucky enough to fraudulently convince a license technician that the code you copied from the sticker should actually be re-bound to new hardware.
Nope, it certainly doesn't. And it will continue to not exist until people actually go try to make it - that's kind of the whole point of trying, isn't it? Nobody will ever develop micro-gravity mining or metallurgy processes here on Earth, they'd be completely useless.
As for doing anything on "industrial scale" - the solution is simple: we won't. Not at first. First we figure out how to do it, and then we figure out how to scale up. Just like we did with terrestrial technology. And that will be fine, because at first we won't have sufficient demand for anything to justify producing it at industrial scales - well, except for rocket fuel maybe.
And a great deal of terrestrial technology will translate with minimal adaptation: casting metal sheets may be a challenge (or not: centrifuges are easy to make in free fall), but I doubt your average roller mill actually depends on gravity for much of anything. Nor does a power-hammer have any special need of gravity. Nor a metal-stamping machine. There will be details where an assembly line will need to be adapted to the fact that gravity isn't a convenient adhesive and sorting tool, but a great deal of the processes could already be migrated with hardly any adaptations at all. Heck, we already have 3D printers that work just fine upside down - microgravity is much less of a challenge.
You've got that a bit wrong there: as you say, the computer is already the brains, AI is good enough. Humans are already the brawn - which means that to replace us, they need automated brawn of comparable utility. And human hands and bodies are impressively nimble and versatile. It's robotics that aren't good enough yet, but they're advancing rapidly.
Nope, but a good impact wrench already does a better job anyway. Just need to make sure you fasten yourself to the car first, so that it doesn't just send you spinning instead. Gravity is really convenient, but it's not irreplaceable.
Of course it still works - nothing has changed except that you can't rely on gravity to hold your feet to the ground. We don't have decent gecko-grip boots yet, but there's plenty of cruder solutions. And for heavier work: power hammers and wood splitters have their own integrated anvils, it's not a complicated concept to adapt. Similarly, if you want to apply forces against an asteroid, you just have to fasten yourself to the ground first so that you don't just push yourself away instead.
Rule of thumb: if you can figure out how to do something upside down or sideways, zero G won't be a problem.
Injection molding doesn't care about gravity, but if you really want to do an open-pour casting, centrifuges are really easy to make in space - all you need is a spinning tether with cages on the ends. And they already use them here on Earth all the time for castings where gravity just doesn't supply as much force as they'd like.
Additional note: consider using rewriteable discs, as their phase-changing crystal medium is far more stable than the organic dye used in write-once discs.
I've lost way too much data to CD/DVD "bit rot" over the years. Multiple copies help, but write-once discs just have lousy shelf lives.
No it wouldn't. At least, not assuming that you downloaded your nice fresh copy of Windows 7 (sans license) free from Microsoft. Microsoft's usage licensing has nothing to do with copyright.
Meanwhile, that license key glued to the case is likely useless for anything other than the motherboard that originally came in that case, since OEM licenses are generally non-transferable to new hardware, though you may luck into a sympathetic license management technician if you call Microsoft.
Nope. Drop your socket, and it slowly starts do drift away from where you let it go. As long as you don't actually throw it away from you, it will remain nearby for quite some time. And since you can't throw fast enough to appreciably change its orbit, the only way it's on a decay orbit to Scotland is if you were already on that path yourself.
Heck, it doesn't even have any differential in gravitational acceleration to power it's rolling underneath to the exact center of your space-car like it inevitably would here on Earth. And if it somehow manages the feat anyway - well, there's no ground underneath it to make it difficult to get to (which would probably dramatically reduce the odds of it happening in the first place).
>we simply do not have the infrastructure in place to mine, purify and process minerals into finished metals in space
We may not need to. There's considerable evidence to suggest that there's plenty of asteroids out there that are nearly pure iron - as in all we have to do is chop it up, hammer it out, or melt it down and cast/print with it. In fact, there was some research recently suggesting that most, if not all, pre iron-age iron tools were made from meteoric iron. Unlike earth-based iron deposits that are all oxides and other iron-bearing minerals, iron meteorites were already fairly pure refined metal: oxidation is primarily an planetary phenomena - you mostly need liquid water or free oxygen to convert iron into non-metallic minerals, and you don't get those in vacuum.
Yeah, such rough-and-ready construction wouldn't benefit from modern advances in alloy technologies - but even cast, rolled, etc. iron is pretty useful stuff, especially in an environment where there's no gravity to resist. Similarly, there's lots of rocky asteroids out there that should provide copious raw filler material for "concretes" of various sorts, which have great potential as radiation shielding. Even if 10% of the material is some sort of vacuum-hardening epoxy bonding agent made on Earth, you can still get 90% of your material from space.
But as you say, not all improvements can be transferred to space. And perhaps far more significantly, a great deal of improvements for space can't be transferred to (or fully developed on) Earth. We've got to actually get out there and start experimenting, see how the theory translates to practice, and have ongoing incentive to improve freefall industrial techniques in order to see the sorts of advances we've seen on Earth.
Yep. And you can download Win10 from MS today, and install it on a PC with a Win7/8 license, so long as you either use assistive technologies or are at least willing to fraudulently claim to. Nothing he's doing changes things in the slightest - except that you don't have to download the several gig.iso file and burn it yourself.
Useless MS software though, which you can download free from Microsoft, since MS has shifted everything to usage licenses with online activation. If you don't already have a license, the software is useless. He was selling the convenience of not having to download and burn it yourself.
Of course, copyright still applies, so useless or not, he would need a license to copy it legally.
Well, that is still copyright infringement though. Just because MS is giving things away for free, doesn't mean that you can do so - it's still protected by copyright law, which means you need an explicit license to duplicate it.
Perhaps you could have an argument if you individually downloaded every copy before burning. Horribly wasteful of course, but then you could argue that Microsoft was the one distributing copies, and you were only redistributing software you had acquired legally. The fact that the final result is identical either way might be enough to give a savvy lawyer an argument to stand on.
Perhaps technically - but he was redistributing something that's already freely available. Sounds like he was basically just doing the "locate, download, and burn DVD" step - no new licenses, no breaking of copy protection, just saving people the trouble of finding, downloading and burning the discs themselves.
Of course the bit about "If I had just written 'Eric's Restore Disc' on there, it would have been fine," suggests that perhaps the real crime is trademark infringement, as the discs were made to resemble the original restore discs, right down to the trademarked logos. Or perhaps he just meant they wouldn't have drawn official attention.
I would assume so as well - it makes wonderful sense to have at least one major wired hub every, say, thousand miles or so, then all satellites will always have a hub within 500 miles, and can do a single-hop customer-to-hub link to the nearest hub.
Where you get a lot of extra flexibility though, is when you realize that 500 miles or horizontal displacement is barely a stretch for a satellite - power is still 1/2 of what it is in the pure vertical case. And even with that limitation you can create single-hop point-to-point links spanning 1000 miles on the surface. So the reality will generally be that each satellite will be in range of several different different hubs at any given moment, as well as having several different satellites capable of covering any particular customer region.
Apply a little whole-system analysis, and it should be relatively easy to arrange for the "hops" to direct traffic to hubs in close geographic/network proximity to the destination. Especially if you assume each satellite can target several sites simultaneously - 2-3 hub links, plus several customer zones, could potentially be routed extremely efficiently. Maintain one link to a hub next door to a Netflix distribution hub, and you can avoid sending all that video across the wired internet entirely, while still utilizing only a single satellite hop.
Yes, there's numbers in the article: 500km orbit. Meaning ~1000km ground-to-ground. Meaning roughly 3-1/3 ms of broadcast latency. Up to twice that for a link between points ~1,700km apart.
I'll admit, it would have been nice if the writers had included such numbers themselves.
Slashdot Mirror
"Actual" or "General" Intelligence is irrelevant. As is "understanding", "comprehension", "self awareness" and any other hurdles you put in the definition of "True AI".
What is relevant, is that we are making machines capable of displaying sophisticated "identify a square" levels of pseudo-"intellectual" achievement. We can now offload new and ever-growing classes of task that used to require human intelligence to a machine, where it can be accelerated and parallelized to a degree completely financially infeasible when relying on humans. The refinement of the mechanical calculator doesn't begin to compare to what we have unlocked.
The robots responsible for doing the majority of trading on the stock market anymore do not understand what they're doing - but that doesn't actually matter. They can still process a far vaster sea of information than a human mind could begin to absorb, and find patterns to guide them faster and more reliably than a human. Usually. It's that failure mode that's one of the other big problems - because they don't actually understand what they're doing, they don't just stumble, they go off the rails at full throttle. And especially in a world where personal profits have long been divorced from personal responsibility, that sort of "win for me, right up until you lose for everybody else" bargain is *incredibly* tempting.
>acquaintance with real facts will, in the end, be better for all parties.
Hogwash. The public's knowledge of the ease and methods of adulterating milk dealt a heavy blow to the profits of unscrupulous milkmen, it was certainly not better for them.
And unlike milk, the potential for abuse of AI is disproportionately focused on the most wealthy and powerful individuals who can afford the hardware necessary to leverage it most effectively. How dare you suggest that such upstanding individuals, the very backbone of our capitalist government, should be subjected to the sort of public scrutiny that common knowledge of the ease with which they might abuse their position would no doubt engender. How dare you sir!
It's not the intelligence of machines directly we need to fear - there doesn't seem to be any immediate threat of true autonomy / free will any time soon. It's the fact that their limited intelligence has no ethical limitations, and can be harnessed to easily enable things that would be prohibitively expensive to do any other way.
As one example: cameras on every street corner - several countries have done that already, and the results are a bit unsettling to anybody who has ever read 1984, but the actual potential for abuse is limited, because there's just too much information to sift through efficiently. Feed all those A/V feeds through an AI that can use facial/gate/etc. recognition and speech analysis to create an easily searchable database of both the real-time and historical movements and public conversations of every person in the country, with anything "suspicious" automatically red-flagged - suddenly 1984 is just an authoritarian administration away.
Or for a more violent dystopia - autonomous weapons platforms, up to and including robotic soldiers/police will allow whoever controls them to crack down on the population just as hard as they want, with none of the (limited) constraints of using human soldiers who may eventually rebel at massacring innocent civilians on a regular basis.
I agree with you about the futility of trying to limit information though - at best you can keep it out of the hands of the general public, but there's no way you're keeping it out of the hands of the wealthy/powerful individuals most likely to abuse it.
True - though it depends very much on what part of the job needs human skills. For example, if you need a human to take a package up the stairs and ring the doorbell for delivery, then it probably doesn't do you much good to automate the delivery truck and render the delivery person into an idle passenger riding between stops.
Well, nothing's going to infect it while it's offline, and if you can't trust your retrieval process, then you likely have bigger problems on your hands. It is nice that write once media protects you from corruption during the retrieval process though - at least assuming you're not using a write-capable drive to read the backups: even if you can't meaningfully update write-once media, it's still generally easy enough to intentionally corrupt it.
I have to agree whole-heartedly about SSDs though - flash stores data as charge-levels on capacitors that start losing charge the moment the data is written. It may be a lot more stable than DRAM, but you couldn't convince me to trust it for archival purposes.
It's easy enough to create an oxygen-free atmosphere here on Earth - and yet we don't work metal that way here either. Perhaps there's a good reason?
First I assume you mean *melting* the iron, rather than boiling it - trying to precisely work with a gas tends to be a laboratory exercise, not an industrial one. Moreover, gaseous iron isn't magnetic, in fact the first magnetic gas was only discovered in 2009, and it was lithium cooled to within a billionth of a degree of absolute zero. And actually, molten iron isn't magnetic either - metals pretty much all lose their magnetic properties well before they melt.
Molten iron is still paramagnetic though, so you could still manipulate it with sufficiently strong magnetic fields. No longer a simple task though, especially considering that whatever iron structure you're building *will* be magnetic, and thus react far more strongly to your magnetic "nozzle" than the molten "ink" does. Plus, if you're spraying liquid metal around it's going to be almost impossible to keep it from splashing, and thus building up on everything else in the area in a chaotic fashion, including your manufacturing equipment.
As for micrometer precision - I think you'd be greatly disappointed. Completely aside from the difficulties of creating such precision, you'd need an extremely narrow stream of metal for than kind of precision to even matter - which means your nozzle has to be very close to the target to keep the stream from solidifying long before it makes contact. Vacuum insulates you from conductive heat losses, but does nothing against radiant losses. So you're basically back to a boring 3D printer, and all the magnetic guidance stuff just introduces needless complexity.
And finally, even assuming all the other problems could be solved, there's one big honking one that probably wont be: material strength. All that hammering and crushing of iron isn't just to get it into a specific shape, it also imparts far greater strength than is present in the original, brittle, cast-iron stock. A very large percentage of the strength and durability of iron products is due to the internal molecular alignments created through working it. Even with today's high-end 3D metal printing, there's usually a final firing process designed to *almost* melt the structure so that the layers merge together at the molecular level, dramatically increasing strength. And even with that, you'd only be looking at solid-cast strength, at best. Worked strength is likely to be permanently out of reach of 3D printing.
Not at all. All it assumes is that making things out of iron would be useful. More sophisticated alloys and manufacturing techniques would undoubtedly be developed too - but we have to start somewhere, and iron is an incredibly useful material just as it is, and far easier to work with when we're just starting out.
As we expand into space, it makes sense to pick the low-hanging fruit first. Bulk construction using iron, radiation shielding using "cemented" dust and gravel, and rocket fuel from water - the technologies for all three are reasonably straightforward, and they address the most mass-intensive demands for developing infrastructure in space.
So where's the disagreement? As I said, the license code off an old PC case sticker is almost certainly going to be useless for anything else. And without a valid license code, so will that shiny new copy of Windows you just downloaded from Microsoft.
No copyright infringement or license violations involved, unless you get lucky enough to fraudulently convince a license technician that the code you copied from the sticker should actually be re-bound to new hardware.
Nope, it certainly doesn't. And it will continue to not exist until people actually go try to make it - that's kind of the whole point of trying, isn't it? Nobody will ever develop micro-gravity mining or metallurgy processes here on Earth, they'd be completely useless.
As for doing anything on "industrial scale" - the solution is simple: we won't. Not at first. First we figure out how to do it, and then we figure out how to scale up. Just like we did with terrestrial technology. And that will be fine, because at first we won't have sufficient demand for anything to justify producing it at industrial scales - well, except for rocket fuel maybe.
And a great deal of terrestrial technology will translate with minimal adaptation: casting metal sheets may be a challenge (or not: centrifuges are easy to make in free fall), but I doubt your average roller mill actually depends on gravity for much of anything. Nor does a power-hammer have any special need of gravity. Nor a metal-stamping machine. There will be details where an assembly line will need to be adapted to the fact that gravity isn't a convenient adhesive and sorting tool, but a great deal of the processes could already be migrated with hardly any adaptations at all. Heck, we already have 3D printers that work just fine upside down - microgravity is much less of a challenge.
You've got that a bit wrong there: as you say, the computer is already the brains, AI is good enough. Humans are already the brawn - which means that to replace us, they need automated brawn of comparable utility. And human hands and bodies are impressively nimble and versatile. It's robotics that aren't good enough yet, but they're advancing rapidly.
Nope, but a good impact wrench already does a better job anyway. Just need to make sure you fasten yourself to the car first, so that it doesn't just send you spinning instead. Gravity is really convenient, but it's not irreplaceable.
Of course it still works - nothing has changed except that you can't rely on gravity to hold your feet to the ground. We don't have decent gecko-grip boots yet, but there's plenty of cruder solutions. And for heavier work: power hammers and wood splitters have their own integrated anvils, it's not a complicated concept to adapt. Similarly, if you want to apply forces against an asteroid, you just have to fasten yourself to the ground first so that you don't just push yourself away instead.
Rule of thumb: if you can figure out how to do something upside down or sideways, zero G won't be a problem.
Injection molding doesn't care about gravity, but if you really want to do an open-pour casting, centrifuges are really easy to make in space - all you need is a spinning tether with cages on the ends. And they already use them here on Earth all the time for castings where gravity just doesn't supply as much force as they'd like.
One alternative I've heard spoken well of is offline hard drives - i.e. pull them out and store them in a cool, dry, safe place. Any thoughts on that?
Additional note: consider using rewriteable discs, as their phase-changing crystal medium is far more stable than the organic dye used in write-once discs.
I've lost way too much data to CD/DVD "bit rot" over the years. Multiple copies help, but write-once discs just have lousy shelf lives.
No it wouldn't. At least, not assuming that you downloaded your nice fresh copy of Windows 7 (sans license) free from Microsoft. Microsoft's usage licensing has nothing to do with copyright.
Meanwhile, that license key glued to the case is likely useless for anything other than the motherboard that originally came in that case, since OEM licenses are generally non-transferable to new hardware, though you may luck into a sympathetic license management technician if you call Microsoft.
Nope. Drop your socket, and it slowly starts do drift away from where you let it go. As long as you don't actually throw it away from you, it will remain nearby for quite some time. And since you can't throw fast enough to appreciably change its orbit, the only way it's on a decay orbit to Scotland is if you were already on that path yourself.
Heck, it doesn't even have any differential in gravitational acceleration to power it's rolling underneath to the exact center of your space-car like it inevitably would here on Earth. And if it somehow manages the feat anyway - well, there's no ground underneath it to make it difficult to get to (which would probably dramatically reduce the odds of it happening in the first place).
>we simply do not have the infrastructure in place to mine, purify and process minerals into finished metals in space
We may not need to. There's considerable evidence to suggest that there's plenty of asteroids out there that are nearly pure iron - as in all we have to do is chop it up, hammer it out, or melt it down and cast/print with it. In fact, there was some research recently suggesting that most, if not all, pre iron-age iron tools were made from meteoric iron. Unlike earth-based iron deposits that are all oxides and other iron-bearing minerals, iron meteorites were already fairly pure refined metal: oxidation is primarily an planetary phenomena - you mostly need liquid water or free oxygen to convert iron into non-metallic minerals, and you don't get those in vacuum.
Yeah, such rough-and-ready construction wouldn't benefit from modern advances in alloy technologies - but even cast, rolled, etc. iron is pretty useful stuff, especially in an environment where there's no gravity to resist. Similarly, there's lots of rocky asteroids out there that should provide copious raw filler material for "concretes" of various sorts, which have great potential as radiation shielding. Even if 10% of the material is some sort of vacuum-hardening epoxy bonding agent made on Earth, you can still get 90% of your material from space.
But as you say, not all improvements can be transferred to space. And perhaps far more significantly, a great deal of improvements for space can't be transferred to (or fully developed on) Earth. We've got to actually get out there and start experimenting, see how the theory translates to practice, and have ongoing incentive to improve freefall industrial techniques in order to see the sorts of advances we've seen on Earth.
That's not surgery, that's just dissection.
Now, if you can put the rock back together and have it work better than before, *then* it's surgery.
Yep. And you can download Win10 from MS today, and install it on a PC with a Win7/8 license, so long as you either use assistive technologies or are at least willing to fraudulently claim to. Nothing he's doing changes things in the slightest - except that you don't have to download the several gig .iso file and burn it yourself.
Useless MS software though, which you can download free from Microsoft, since MS has shifted everything to usage licenses with online activation. If you don't already have a license, the software is useless. He was selling the convenience of not having to download and burn it yourself.
Of course, copyright still applies, so useless or not, he would need a license to copy it legally.
Well, that is still copyright infringement though. Just because MS is giving things away for free, doesn't mean that you can do so - it's still protected by copyright law, which means you need an explicit license to duplicate it.
Perhaps you could have an argument if you individually downloaded every copy before burning. Horribly wasteful of course, but then you could argue that Microsoft was the one distributing copies, and you were only redistributing software you had acquired legally. The fact that the final result is identical either way might be enough to give a savvy lawyer an argument to stand on.
Perhaps technically - but he was redistributing something that's already freely available. Sounds like he was basically just doing the "locate, download, and burn DVD" step - no new licenses, no breaking of copy protection, just saving people the trouble of finding, downloading and burning the discs themselves.
Of course the bit about "If I had just written 'Eric's Restore Disc' on there, it would have been fine," suggests that perhaps the real crime is trademark infringement, as the discs were made to resemble the original restore discs, right down to the trademarked logos. Or perhaps he just meant they wouldn't have drawn official attention.
I would assume so as well - it makes wonderful sense to have at least one major wired hub every, say, thousand miles or so, then all satellites will always have a hub within 500 miles, and can do a single-hop customer-to-hub link to the nearest hub.
Where you get a lot of extra flexibility though, is when you realize that 500 miles or horizontal displacement is barely a stretch for a satellite - power is still 1/2 of what it is in the pure vertical case. And even with that limitation you can create single-hop point-to-point links spanning 1000 miles on the surface. So the reality will generally be that each satellite will be in range of several different different hubs at any given moment, as well as having several different satellites capable of covering any particular customer region.
Apply a little whole-system analysis, and it should be relatively easy to arrange for the "hops" to direct traffic to hubs in close geographic/network proximity to the destination. Especially if you assume each satellite can target several sites simultaneously - 2-3 hub links, plus several customer zones, could potentially be routed extremely efficiently. Maintain one link to a hub next door to a Netflix distribution hub, and you can avoid sending all that video across the wired internet entirely, while still utilizing only a single satellite hop.
Yes, there's numbers in the article: 500km orbit. Meaning ~1000km ground-to-ground. Meaning roughly 3-1/3 ms of broadcast latency. Up to twice that for a link between points ~1,700km apart.
I'll admit, it would have been nice if the writers had included such numbers themselves.