Actually "Son of SpaceShip Two", otherwise known as Stratolaunch ( http://www.stratolaunch.com/) , is the special thing. It's the same basic idea, carrier airplane and rocket stages, but much bigger and able to put significant payloads in orbit. Yet another billionaire, Microsoft co-founder Paul Allen, started this up. The same people (Scaled Composites) are building this as SpaceShip Two. But this time they are using parts from *two* 747's to build the world's largest airplane, to maximize launch capacity.
At first, the rocket stages are disposable, and only the airplane comes back, but in the long run I expect them to reuse the rocket stages too, it would save a lot per flight.
> Yeah, I buy it. Hell, I'd work for little more than a pretty meager wage if I could be reasonably sure of ACCOMPLISHING something meaningful in space.
You can, and don't even have to quit your day job. Read up on "self expanding automation" ( http://en.wikibooks.org/wiki/S... ), where a starter kit of machines is used to build parts for more machines, until you have the production capacity you need. In space, instead of sending a whole factory to process asteroids or support a Mars base, you send just the starter kit, and build the rest from local materials. On Earth the same idea of bootstrapping from a starter kit lets you grow your rocket factory or launch site, or any other industrial production you want.
An industrial starter kit will be beyond the average person's finances, but then so is a data center. The way to bring it down to the individual level is (a) to make it distributed - the machines are split up among different people or groups, but they collaborate to make things, or (b) to run a centralized production location on a time share or lease basis. You need a part, you submit the design files and a small payment to have it made. Much of the work becomes designing new or upgraded things to make to expand the system. You can do that at home, because you already have a computer to run the design software.
Once people get the hang of making starter kits and growing them, they can work towards more and more difficult locations. First typical metro areas, then deserts, ice caps, oceans, and finally space. Your first kit, when it is fully grown up, can produce another starter kit for the next location, and so on. You automatically build your supply chain as you go.
What was the value of the Nevada desert before we built Las Vegas? What's it worth now? Musk is the equivalent of the Connestoga wagon builders for the first settlers. Real estate developers will do the bulk of the work, like they do on Earth. Remember that Mars has the same land area as the Earth (minus oceans), and none of it is claimed yet.
It will if you do it right. Self-expaning factories and mining everywhere would bring the cost down to that range.
* Self-expanding production is where you deliver a "starter kit" of machines. Those machines make parts for more machines out of local materials. An example is using CNC machine tools to make metal parts out of Metallic type asteroid rock. The expanding collection of machines then can make more and more items.
* Sourcing your raw materials locally, in every orbit, turns the exponential rocket equation into a linear problem.
The combination of the two has the potential to reduce the cost of supporting a human on Mars by a factor of 2500 over today's cost. You won't get that reduction all at once, but it brings the cost of a colony (mostly self-supporting, lots of people) within reach.
> . Asteroid mining isn't going to do anything, directly, for people on Earth.
This is incorrect. Space is already a $300 billion/year industry, of which NASA only represents 6%. Most of it is in the thousand or so satellites in Earth orbit, doing communications, navigation, weather, mapping, etc. Today, when a satellite breaks a part or runs out of fuel, there is no way to fix or refuel it. So you have to launch a whole new satellite at great expense. An "orbital service station" with the capacity to do those things is worth billions a year. Since Near Earth Asteroids can supply 30-50 times more fuel, plus the potential for other supplies, it makes that service station cheaper to operate. That's enough of a "first market" to bootstrap the development. Once you are up and operating, other activities become economic besides satellite servicing.
> We've spent the last half century turning into a knowledge society
And all of those billions of computers run on energy, as does the rest of civilization. Modern civilization is based on replacing human and animal labor with mechanical and electrical power. The amount of solar energy passing closer than the Moon is equal to the whole world's fossil fuel reserves *every minute*. We just have to learn how to exploit it. Leveraging resources already on location is part of that equation.
Space industry worldwide is $300 billion a year, of which NASA is about 6%. Most of the money, and most of the recent technology improvements, are from satellite communications. High efficiency solar arrays and ion propulsion have been used on satellites for about 15 years now. The Dawn asteroid mission (which has electric thrusters) was 7 years later and 1/4 the mass.
The Rocket Equation gives us guidance on how to get around it. Increase exhaust velocity or reduce velocity increments. For lots more detail, see my book ( http://en.wikibooks.org/wiki/S... ), but here are a few ideas:
* Replace some of the bottom part of reaching orbit with higher efficiency engines. That can be anything from subsonic jets (Stratolaunch) to ramjets, to ground accelerators. Replace some of the top part of reaching orbit with electric thrusters transferring momentum to a fractional space elevator (only reaches part way to the ground. By reducing the velocity provided by chemical engines, and more by other methods, you lower the mass ratio.
* Mine for fuel everywhere: scoop mine our atmosphere from orbit, process asteroid rock in high orbit, mine Phobos for fuel, etc. By repeatedly fueling at each location, you reduce fuel needed to a series of linear steps, instead of an exponential. If fuel extraction has a large mass return on your capital equipment, how much you need to launch from Earth drops dramatically.
> Whereas most people in slashdot are brainwashed and gullible morons that would rather invest in the most useless and most expensive form of energy generation: photovoltaics.
I guess you think Warren Buffet is a moronic investor too:
> It reminds me of that old joke about why the Space Shuttle (and now SLS) design is influenced by the width of a horses ass.
Two horse's asses. Wagons were sized according to the width of two horses, and roads were in turn sized to fit the wagons. When underground mining got serious, the mining wagons were just converted outdoor wagons, still pulled by two horses. Then they started putting rails under the wagons, to allow moving heavier loads with less friction. Rail lines began to be used outdoors, pulled by horses at first, so the rail spacing continued to be suitable to the width of two horses. One engines replaced horses, the rails stayed the same width. Go look at train tracks today, you will see they are the right size for two horses to fit.
The Solid rocket boosters for the Space Shuttle were shipped by rail from Utah to Florida, and thus had to fit on railcars on the standard rail spacing. In turn, the size of the boosters set the lift capacity of the rocket, and thus how big the Shuttle Orbiters could be. Finally, the Space Station modules had to fit in the Orbiter, so the Space Station's design is dictated by the width of two horse's asses.
I may have been responsible for this analysis about 30 years ago at Boeing. I was both designing launch vehicles, and had a hobby interest in the history of technology. It is also possible it came up in a USENET discussion on sci.space back then. I don't remember any more.
Since this was the post office calling, the answer would be to go to the local machine shop and borrow one for a minute. Since they deliver to everyone in town anyway, just assign whichever letter carrier has that shop on their route to do it.
> Automation will destroy jobs to the extent that the people running the companies implementing the automation wish it to.
I disagree with this statement, and the assumptions behind it. First is the assumption that only big companies can have automation. That like assuming only big companies can have computers. 50 years ago that was a reasonable assumption, because computers were expensive, like industrial robots are today. But it need not be true in the future. I'm working on "MakerNets" and "Seed Factories". A MakerNet is a network of people with skills and some machines. They help each other build stuff and upgrade. Eventually you reach a level where most of the member's needs are met by automation within the network. If your housing, food, and utilites are supplied that way, most of the need for conventional jobs goes away.
A "Seed Factory" is a starter kit of machines designed to make parts for more machines to expand the factory capacity. At first, the starter kit won't make anywhere close to 100% of the new parts. Stuff will still have to be bought. But as the collection of machines grows, you can make more of your own items, and need to buy less. The engineering question is what belongs in the starter kit, and what's the best path for expansion.
The problem with corporations is the separation of ownership and labor, giving them conflicting goals. Concepts like MakerNets and Seed Factories lets the workers "grow their own" production. Then they are owners as well as workers for whatever amount of labor is still needed. Their goals are now aligned - owner-operators are not going to lay themselves off. More automation just makes them more efficient.
To a first approximation, everyone in the world has a mobile phone, and the percentage that have a smartphone is rapidly increasing. Everyone will be able to *access* an AI, just like we can access Google search. Your comment is like "the poor won't be able to afford a library", it is poorly formulated. You don't need your own full time AI, just enough access to do the things you need to do.
As a space systems engineer, I would agree. There isn't much benefit to putting humans in the Venusian clouds. It adds weight and risk to the mission. They can just as well stay in orbit, controlling a robotic exploration blimp.
Essentially *all* Dollars, in a computer or in paper, are just representations of debt. When the Federal Reserve issues paper money, it's backed by Treasury bonds and mortgages. When banks make loans, the loan check is backed by the loan they just made.
> I don't believe that it is absolutely uncountefeitable. I did read the Wiki on it, but I don't think that any item digitally-based is safe. We are on/. after all;)
Then you don't understand the central invention in bitcoin, an accounting ledger that is public and cannot be altered. In order to counterfeit some bitcoins, you would have insert a transaction into everyone's copy of the ledger, and also have a valid checksum (hash). Without it, every copy of the software will reject the data as invalid. The huge amount of specialized hardware thrown at generating valid hashes prevents anyone else from inserting a spurious one.
> I just hope it does not become publicly accepted. Can someone help convince me that bitcoin is the way to go?
Modern money, like the US Dollar, is just how we keep track of who owes something to others, or is owed something. Mostly the tracking is via bank ledgers, but occasionally it is on paper (as in paper money). Almost every dollar in existence represents a debt, either treasury bonds, mortgages, or other bank loans. Dollars make the debt easily traded in convenient units.
Bitcoin also uses a ledger to keep track. It's a public distributed one, but rather than being based on debt, it's based on work. All the balances are positive. This discourages a government from spending money it doesn't have, and banks from leeching off the rest of us, earning money simply because they create debt, instead of actual work.
Central banks have shown time and time again that they are bad at managing their money supply. Venezuela and Russia are two current examples, but there are literally hundreds of examples through history. Bitcoin's supply is managed by a software algorithm that releases new units at a known rate. It is on automatic pilot rather than the whim of fallible humans. I trust that more.
The big drop since 2008 is because solar cell production at that point exceeded silicon ingot needs for electronics. When electronics was the main demand, it was not worth making separate plants for "solar grade" silicon. Electronics grade is much more expensive, because crystal defects make computer chips unusable. They only make solar cells a little less efficient. Nowadays silicon ingots for solar panels is the dominant market. In this graph you can even see a rise in price as silicon got in short supply, followed by the crash as new solar-grade production went into operation:
Nope. They require a considerable amount of energy to create, that of all the mining rigs combined. The energy is used to update the transaction history (block chain). One of the transactions adds 25 new bitcoins to the history, but that event can't happen without finding a hash value for the block. Finding the hash value requires all the mining rigs to search for it, using energy.
In the context of energy supplies, renewable means "there will be more supply tomorrow", i.e. the supply constantly renews itself. New coal and oil are not being produced at any appreciable rate. However the Sun is still fusing, and rain will refill dam reservoirs.
From a physics standpoint, yes, the Sun will eventually run out of fuel, but that's meaningless from a human standpoint. It will last a million times longer than we have had civilization.
> Maybe you hadn't heard, but there are people being paid to work out how to do all these things,
Yup, I'm one of them. I'm working on the idea of a "Seed Factory" ( http://en.wikibooks.org/wiki/S... ), which is a starter kit of machines which you expand by making parts for more machines. A 3D printer is likely to be part of the starter kit, but you need several others. The engineering R&D questions are what machines should be in the starter kit, what is the optimal growth path, and how do those vary with available raw materials and the desired end outputs.
The point here is sending a whole industrial plant into orbit to process an asteroid is much too massive. You want to bootstrap up from a minimal starter kit, and build the rest out of the asteroid itself, as much as possible. You won't reach 100%, some stuff will still need to come from Earth, but saving 85-98% of the launch weight (what we think is a reasonable goal) is still a huge advantage.
Today you need to launch a spool of ABS plastic, but one spool is less mass than a spare of every possible part that can break. It reduces spares inventory. This is also a very small first step towards making everything you need in orbit. ABS plastic is mostly Carbon and Hydrogen, which are both present in Chondrite type asteroids. The remainder is Nitrogen, which can be scoop-mined from the upper atmosphere. Combined you have all the ingredients for the plastic. You need other stuff in larger quantities in orbit - fuel, radiation shielding, oxygen, water, etc, so those will be produced first.
Now that they have a proof of concept, it is an obvious thing for researchers to try different pit sizes and patterns in order to optimize the efficiency. One thing they probably haven't checked yet is the effect of Sun angle. Most solar panels are on a fixed mounting. So the Sun lights them from different angles during the day. Therefore any patterning will have a different apparent pit spacing. I think there is a lot more to learn about this effect, but even small efficiency gains have dramatic effects. Most of the cost of a solar system these days is the "balance of system", i.e. everything besides the panels. More efficient panels means you need less of them, and therefore less installation labor and other overhead costs.
Slashdot Mirror
> VG won't be doing anything special,
Actually "Son of SpaceShip Two", otherwise known as Stratolaunch ( http://www.stratolaunch.com/) , is the special thing. It's the same basic idea, carrier airplane and rocket stages, but much bigger and able to put significant payloads in orbit. Yet another billionaire, Microsoft co-founder Paul Allen, started this up. The same people (Scaled Composites) are building this as SpaceShip Two. But this time they are using parts from *two* 747's to build the world's largest airplane, to maximize launch capacity.
At first, the rocket stages are disposable, and only the airplane comes back, but in the long run I expect them to reuse the rocket stages too, it would save a lot per flight.
> Yeah, I buy it. Hell, I'd work for little more than a pretty meager wage if I could be reasonably sure of ACCOMPLISHING something meaningful in space.
You can, and don't even have to quit your day job. Read up on "self expanding automation" ( http://en.wikibooks.org/wiki/S... ), where a starter kit of machines is used to build parts for more machines, until you have the production capacity you need. In space, instead of sending a whole factory to process asteroids or support a Mars base, you send just the starter kit, and build the rest from local materials. On Earth the same idea of bootstrapping from a starter kit lets you grow your rocket factory or launch site, or any other industrial production you want.
An industrial starter kit will be beyond the average person's finances, but then so is a data center. The way to bring it down to the individual level is (a) to make it distributed - the machines are split up among different people or groups, but they collaborate to make things, or (b) to run a centralized production location on a time share or lease basis. You need a part, you submit the design files and a small payment to have it made. Much of the work becomes designing new or upgraded things to make to expand the system. You can do that at home, because you already have a computer to run the design software.
Once people get the hang of making starter kits and growing them, they can work towards more and more difficult locations. First typical metro areas, then deserts, ice caps, oceans, and finally space. Your first kit, when it is fully grown up, can produce another starter kit for the next location, and so on. You automatically build your supply chain as you go.
What was the value of the Nevada desert before we built Las Vegas? What's it worth now? Musk is the equivalent of the Connestoga wagon builders for the first settlers. Real estate developers will do the bulk of the work, like they do on Earth. Remember that Mars has the same land area as the Earth (minus oceans), and none of it is claimed yet.
It will if you do it right. Self-expaning factories and mining everywhere would bring the cost down to that range.
* Self-expanding production is where you deliver a "starter kit" of machines. Those machines make parts for more machines out of local materials. An example is using CNC machine tools to make metal parts out of Metallic type asteroid rock. The expanding collection of machines then can make more and more items.
* Sourcing your raw materials locally, in every orbit, turns the exponential rocket equation into a linear problem.
The combination of the two has the potential to reduce the cost of supporting a human on Mars by a factor of 2500 over today's cost. You won't get that reduction all at once, but it brings the cost of a colony (mostly self-supporting, lots of people) within reach.
> . Asteroid mining isn't going to do anything, directly, for people on Earth.
This is incorrect. Space is already a $300 billion/year industry, of which NASA only represents 6%. Most of it is in the thousand or so satellites in Earth orbit, doing communications, navigation, weather, mapping, etc. Today, when a satellite breaks a part or runs out of fuel, there is no way to fix or refuel it. So you have to launch a whole new satellite at great expense. An "orbital service station" with the capacity to do those things is worth billions a year. Since Near Earth Asteroids can supply 30-50 times more fuel, plus the potential for other supplies, it makes that service station cheaper to operate. That's enough of a "first market" to bootstrap the development. Once you are up and operating, other activities become economic besides satellite servicing.
> We've spent the last half century turning into a knowledge society
And all of those billions of computers run on energy, as does the rest of civilization. Modern civilization is based on replacing human and animal labor with mechanical and electrical power. The amount of solar energy passing closer than the Moon is equal to the whole world's fossil fuel reserves *every minute*. We just have to learn how to exploit it. Leveraging resources already on location is part of that equation.
> while space travel is largely a money pit.
Space industry worldwide is $300 billion a year, of which NASA is about 6%. Most of the money, and most of the recent technology improvements, are from satellite communications. High efficiency solar arrays and ion propulsion have been used on satellites for about 15 years now. The Dawn asteroid mission (which has electric thrusters) was 7 years later and 1/4 the mass.
The Rocket Equation gives us guidance on how to get around it. Increase exhaust velocity or reduce velocity increments. For lots more detail, see my book ( http://en.wikibooks.org/wiki/S... ), but here are a few ideas:
* Replace some of the bottom part of reaching orbit with higher efficiency engines. That can be anything from subsonic jets (Stratolaunch) to ramjets, to ground accelerators. Replace some of the top part of reaching orbit with electric thrusters transferring momentum to a fractional space elevator (only reaches part way to the ground. By reducing the velocity provided by chemical engines, and more by other methods, you lower the mass ratio.
* Mine for fuel everywhere: scoop mine our atmosphere from orbit, process asteroid rock in high orbit, mine Phobos for fuel, etc. By repeatedly fueling at each location, you reduce fuel needed to a series of linear steps, instead of an exponential. If fuel extraction has a large mass return on your capital equipment, how much you need to launch from Earth drops dramatically.
> Whereas most people in slashdot are brainwashed and gullible morons that would rather invest in the most useless and most expensive form of energy generation: photovoltaics.
I guess you think Warren Buffet is a moronic investor too:
http://www.bloomberg.com/news/...
> It reminds me of that old joke about why the Space Shuttle (and now SLS) design is influenced by the width of a horses ass.
Two horse's asses. Wagons were sized according to the width of two horses, and roads were in turn sized to fit the wagons. When underground mining got serious, the mining wagons were just converted outdoor wagons, still pulled by two horses. Then they started putting rails under the wagons, to allow moving heavier loads with less friction. Rail lines began to be used outdoors, pulled by horses at first, so the rail spacing continued to be suitable to the width of two horses. One engines replaced horses, the rails stayed the same width. Go look at train tracks today, you will see they are the right size for two horses to fit.
The Solid rocket boosters for the Space Shuttle were shipped by rail from Utah to Florida, and thus had to fit on railcars on the standard rail spacing. In turn, the size of the boosters set the lift capacity of the rocket, and thus how big the Shuttle Orbiters could be. Finally, the Space Station modules had to fit in the Orbiter, so the Space Station's design is dictated by the width of two horse's asses.
I may have been responsible for this analysis about 30 years ago at Boeing. I was both designing launch vehicles, and had a hobby interest in the history of technology. It is also possible it came up in a USENET discussion on sci.space back then. I don't remember any more.
Since this was the post office calling, the answer would be to go to the local machine shop and borrow one for a minute. Since they deliver to everyone in town anyway, just assign whichever letter carrier has that shop on their route to do it.
> Automation will destroy jobs to the extent that the people running the companies implementing the automation wish it to.
I disagree with this statement, and the assumptions behind it. First is the assumption that only big companies can have automation. That like assuming only big companies can have computers. 50 years ago that was a reasonable assumption, because computers were expensive, like industrial robots are today. But it need not be true in the future. I'm working on "MakerNets" and "Seed Factories". A MakerNet is a network of people with skills and some machines. They help each other build stuff and upgrade. Eventually you reach a level where most of the member's needs are met by automation within the network. If your housing, food, and utilites are supplied that way, most of the need for conventional jobs goes away.
A "Seed Factory" is a starter kit of machines designed to make parts for more machines to expand the factory capacity. At first, the starter kit won't make anywhere close to 100% of the new parts. Stuff will still have to be bought. But as the collection of machines grows, you can make more of your own items, and need to buy less. The engineering question is what belongs in the starter kit, and what's the best path for expansion.
The problem with corporations is the separation of ownership and labor, giving them conflicting goals. Concepts like MakerNets and Seed Factories lets the workers "grow their own" production. Then they are owners as well as workers for whatever amount of labor is still needed. Their goals are now aligned - owner-operators are not going to lay themselves off. More automation just makes them more efficient.
To a first approximation, everyone in the world has a mobile phone, and the percentage that have a smartphone is rapidly increasing. Everyone will be able to *access* an AI, just like we can access Google search. Your comment is like "the poor won't be able to afford a library", it is poorly formulated. You don't need your own full time AI, just enough access to do the things you need to do.
As a space systems engineer, I would agree. There isn't much benefit to putting humans in the Venusian clouds. It adds weight and risk to the mission. They can just as well stay in orbit, controlling a robotic exploration blimp.
> carrying out a crime while armed makes the penalty more severe.
Police are always armed when on duty. Just apply the same rules to them as to everyone else.
Essentially *all* Dollars, in a computer or in paper, are just representations of debt. When the Federal Reserve issues paper money, it's backed by Treasury bonds and mortgages. When banks make loans, the loan check is backed by the loan they just made.
> I don't believe that it is absolutely uncountefeitable. I did read the Wiki on it, but I don't think that any item digitally-based is safe. We are on /. after all ;)
Then you don't understand the central invention in bitcoin, an accounting ledger that is public and cannot be altered. In order to counterfeit some bitcoins, you would have insert a transaction into everyone's copy of the ledger, and also have a valid checksum (hash). Without it, every copy of the software will reject the data as invalid. The huge amount of specialized hardware thrown at generating valid hashes prevents anyone else from inserting a spurious one.
> I just hope it does not become publicly accepted. Can someone help convince me that bitcoin is the way to go?
Modern money, like the US Dollar, is just how we keep track of who owes something to others, or is owed something. Mostly the tracking is via bank ledgers, but occasionally it is on paper (as in paper money). Almost every dollar in existence represents a debt, either treasury bonds, mortgages, or other bank loans. Dollars make the debt easily traded in convenient units.
Bitcoin also uses a ledger to keep track. It's a public distributed one, but rather than being based on debt, it's based on work. All the balances are positive. This discourages a government from spending money it doesn't have, and banks from leeching off the rest of us, earning money simply because they create debt, instead of actual work.
Central banks have shown time and time again that they are bad at managing their money supply. Venezuela and Russia are two current examples, but there are literally hundreds of examples through history. Bitcoin's supply is managed by a software algorithm that releases new units at a known rate. It is on automatic pilot rather than the whim of fallible humans. I trust that more.
The big drop since 2008 is because solar cell production at that point exceeded silicon ingot needs for electronics. When electronics was the main demand, it was not worth making separate plants for "solar grade" silicon. Electronics grade is much more expensive, because crystal defects make computer chips unusable. They only make solar cells a little less efficient. Nowadays silicon ingots for solar panels is the dominant market. In this graph you can even see a rise in price as silicon got in short supply, followed by the crash as new solar-grade production went into operation:
http://costofsolar.com/managem...
Nope. They require a considerable amount of energy to create, that of all the mining rigs combined. The energy is used to update the transaction history (block chain). One of the transactions adds 25 new bitcoins to the history, but that event can't happen without finding a hash value for the block. Finding the hash value requires all the mining rigs to search for it, using energy.
> It seems more and more investors are losing faith in the very possibility of producing energy with fusion.
[citation needed]
This recent survey of alternate fusion projects says otherwise: http://nextbigfuture.com/2014/...
In the context of energy supplies, renewable means "there will be more supply tomorrow", i.e. the supply constantly renews itself. New coal and oil are not being produced at any appreciable rate. However the Sun is still fusing, and rain will refill dam reservoirs.
From a physics standpoint, yes, the Sun will eventually run out of fuel, but that's meaningless from a human standpoint. It will last a million times longer than we have had civilization.
> Maybe you hadn't heard, but there are people being paid to work out how to do all these things,
Yup, I'm one of them. I'm working on the idea of a "Seed Factory" ( http://en.wikibooks.org/wiki/S... ), which is a starter kit of machines which you expand by making parts for more machines. A 3D printer is likely to be part of the starter kit, but you need several others. The engineering R&D questions are what machines should be in the starter kit, what is the optimal growth path, and how do those vary with available raw materials and the desired end outputs.
The point here is sending a whole industrial plant into orbit to process an asteroid is much too massive. You want to bootstrap up from a minimal starter kit, and build the rest out of the asteroid itself, as much as possible. You won't reach 100%, some stuff will still need to come from Earth, but saving 85-98% of the launch weight (what we think is a reasonable goal) is still a huge advantage.
Today you need to launch a spool of ABS plastic, but one spool is less mass than a spare of every possible part that can break. It reduces spares inventory. This is also a very small first step towards making everything you need in orbit. ABS plastic is mostly Carbon and Hydrogen, which are both present in Chondrite type asteroids. The remainder is Nitrogen, which can be scoop-mined from the upper atmosphere. Combined you have all the ingredients for the plastic. You need other stuff in larger quantities in orbit - fuel, radiation shielding, oxygen, water, etc, so those will be produced first.
Now that they have a proof of concept, it is an obvious thing for researchers to try different pit sizes and patterns in order to optimize the efficiency. One thing they probably haven't checked yet is the effect of Sun angle. Most solar panels are on a fixed mounting. So the Sun lights them from different angles during the day. Therefore any patterning will have a different apparent pit spacing. I think there is a lot more to learn about this effect, but even small efficiency gains have dramatic effects. Most of the cost of a solar system these days is the "balance of system", i.e. everything besides the panels. More efficient panels means you need less of them, and therefore less installation labor and other overhead costs.