It won't be able to accurately predict how biology will react to a new material
Except that the paper shows that it does. Not 100% accuracy, but 78-96%, quite good enough to do compound screening.
What sorts of "new material" do you think they will encounter? Chemicals made out of new elements?
Chemicals are put together is well defined ways, based on the small number of types of bonds that exist, though combinatorics leads to very large numbers of structures. Most organic molecules are made of just four types of atoms. A few others (phosphorus and sulfur, for example) are fairly common too. All chemicals share motifs, functional groups, and structural features with other chemicals. There are no "completely new" molecules.
(though it might be able to make a statistically better-than-chance guess).
It might indeed. In fact, it does better than standard non-chance prediction methods, and astronomically better than "chance" guesses. That is the whole point of the paper.
That's not machine learning. It's statistical analysis on a large scale.
And that is exactly what machine learning is. Discovering relationships in data patterns on very large sets of data with many attributes using statistical properties.
Molecules of interest generally have many parts ("functional groups") and many structural patterns. So many in fact that coming up with a consistent way of simply naming the molecules has been a challenge for chemistry for generations. And each of these groups and structural components interact with biological systems in different ways. After many decades of animal testing we have a very large set of data about the effects of millions of molecules on animals. This program finds structural pattern matches within this data that predict biological effects, and matches them to new molecules. If that isn't ML there is no ML.
It does not do away with animal testing. One needs to know the effects. Without animal testing you have no understanding of what the chemical will do. The only reason you would not need to perform animal testing is if testing has already been done and the new chemical has similar properties to previously tested chemicals.
It allows up use the vast database of animal tests that have already been done over a century or more. Every new chemical has "similar properties" to other chemicals, usually many chemicals - and many differences too. Finding how those similarities and differences relate to biological effects is what this program does.
The headline and article writer(s) for Phys.org is/are using the term "color" to mean pigment. The should have used "pigment" like the actual researchers do. The summary author here was just copying the Phys.org headline (mis)usage.
If they wanted to stay with "color" they should have used a compound term like "organic color", "biogenic color", or some-such.
All colors are real. Colors are the perceptions of the human visual system.
Appealing to "purple" as a special case of "color" is not well chosen since what is denoted "purple" is ambiguous (actually a problem with color names in general, when viewed cross-culturally). "Brown" is a much better example since it is necessarily and always a color mixture.
You are claiming I think that only spectral colors are real, which they are, but not exclusively. Color mixtures are still colors. Spectral colors are rarely seen under natural conditions (rainbows, and other occasions of color dispersion through refraction).
Things started to get better with the advent of the boost libraries and the C++11 standard brought a lot of the ideas from those libraries over in the standard.
Boost only came out about the start of 2000, at which point C++ had been under development by Stroustrup for 21 years, and the release of C++ 2.0, arguably the first "modern" version of C++ was 11 year earlier. And C++11 was only approved on 12 August 2011.
That is a lot pain and suffering and legacy code before the "better" arrived.
An odd lot, not my cup of tea, but I am scarcely in any position to judge. It is against-the-grain subcultures that help leaven the dough of a healthy society. And indeed the desire no to be surveilled is at the very heart of all liberty.
He3 makes things better, but I don't see how it helps enough.
Really it makes things far worse.
The only area where it helps is reducing the neutron damage and activation of the inner reactor parts, which are estimated to run only 5-8% of the capital contribution to the cost of electricity. But the reaction itself is ten thousand times harder to do (D/D fusion is only a few hundred times harder). We have good ideas for doing D/T fusion, that should work (at unaffordable cost) in several decades. We have none for He-3/D fusion, at this point it while the reaction is real, the technology is wishing for pink unicorns.
And then there is the fuel cost. One kilogram of D can be bought today for $3000. To get the equivalent amount of He-3 from the Moon you have to process ~300,000 tons of regolith on the Moon, and ship it back to Earth. Show me any sort of conceptual process that can do this for a penny a ton. Here on Earth currently total ore extraction and processing costs are in the range of $2 to $200 per ton, it is not going to be 100 to 10,000 times cheaper on the Moon, rather we can expect the reverse to be true.
Mapping the helium-3 distribution on the Moon is a worthy scientific endeavor - it will tell us much about how the solar wind interacts with the lunar surface.
But promoting the project for its "nuclear fuel" potential is so out of line with reality that it is deception, pure and simple.
First there is no prospect of building a helium-3 reactor. We currently cannot build a power-producing fusion reactor using the easiest fuel, deuterium-tritium, even though is reaction rate is ten thousand times faster than He-3/D at plausible temperatures.
Second we already can accurately forecast that when we can build a fusion reactor that uses that easiest to burn D-T fuel it will not be able to compete with any commercial source of electricity. The capital and operating costs of such a plant place the electricity cost at about ten times what wholesale electricity has been selling at for decades (an inflation adjusted current $30/MWh). This recent paper (accessible through Sci-hub) places the economics of a D-T plant in the best possible light and comes up with electricity costs due to the high capital cost of $175-$312 MWh*. Remember that He-3 fusion is ten thousand times harder, and we now have to mine the fuel on the Moon.
The only theoretical advantage of He-3 fusion is the lack of neutron emission from the main reaction (side reactions would still produce some). This would greatly reduce the neutron damage that requires periodic replacement of parts in D/T (or D/D) reactor, and greatly reduce the radioactive waste produced from neutron activated components. These are not major contributors to the projected cost of fusion power (the paper above assigns $14/MWh for these combined, 5-8% of the projected costs), so greatly reducing them does little to improve it.
And long before we can build a working He-3/D reactor, we will be able to build a D/D reactor using cheap, plentiful deuterium, available for a few thousand dollars a kilogram on Earth in effectively unlimited supply. The D/D reaction is "only" a few hundred times harder than D/T.
*The paper ultimately claims that it would be competitive, when externalities are costed, mostly by assigning very high externality costs to every other form of power, and assumes that all of that will be some day captured in electricity pricing. Its treatment of on-shore wind, and solar PV is especially suspect since it assigns levelized costs per MWh, 40 years in the future, that are several times higher than current, demonstrated costs now. This is a lot of special pleading.
Nukes can work for cement, which just needs heat for the kiln. But nuclear aircraft? I don't think so. An iron blast furnace uses metallurgical coal (converted to coke), as an integral part of the process. You can't just drop in nuclear as a replacement.
The emission of CO2 from cement making is more fundamental to the process than it is for iron making.
Cement releases CO2 from the most fundamental chemical reaction required to make it: converting limestone to lime. This is the reaction: CaCO3 -> CaO + CO2. The release of carbon dioxide is unavoidable.
Iron on the other can in principle be made electrolytically although the process needs more development before it can be commercialized.
About half of the CO2 released gets recaptured by the cement over the course of several decades as the cement completes its curing process, but the other half never is. If we can capture CO2 at the cement plant source, that slow curing process might actually be a way to remove CO2 from the air.
Of course if we can capture cement plant CO2 at the source, we can do that with blast furnaces too, so that provides another option.
I am not a "sports fan" of any description, but to claim that "sports don't matter" is only valid if you also make the claim that "entertainment of any form doesn't matter".
I discovered a year or so back, using the iOS Google Maps for driving info, that they aren't actual maps - like you find on paper traffic maps - but very map-looking schematics that omit things they think you don't need to know about.
Like highway on-off ramps.
When you use the app for directions, where it knows your destination, yes, it shows you where to get off (if you follow its directions), but otherwise these traffic connections are not shown. I found this out since I wanted to see what the next off-ramp was, not having a specific destination programmed in. And I found none.
The Apple Maps app however did show every on-off ramp, like a real traffic map.
The true cosmically significant number to express ultimate enthusiasm is 42!!!!!!!!!!!!!!!!!!!!!!!!
And to anyone who is wondering, the enthusiasm limit imposed by/. is 24 (the number you see above). Try 25 and you get "filter error, try using fewer junk characters".
I'll bet you are expressing frustration at being denied the use of the Interrobang
I ended my sentence above with a string of Interrobangs, and they type in the editing window, but do not show in the preview. I will bet that/. is part of the anti-Interrobang conspiracy.
After driving through the Great Plains this past year, where small towns (and I mean small towns) are scattered over vast areas, and where there are many small towns that got so small they closed up shop, and hearing about the problems of getting goods delivered out there, I thought that a good strategy for a behemoth like Amazon that wants to Sell Everything to Everyone would be to set up a system of delivery hubs across the plains.
Buy up abandoned buildings in abandoned (or nearly so) towns, and convert them into the equivalent of post office boxes for packages, with locking storage. With web enabled smart locks the locals would not need dedicated containers, but would be notified which container had their stuff, and they would be able to unlock it to retrieve it.
An additional iteration would be along the lines of the TFA, allowing other locals to sign up for delivering to their neighbors.
Naphthalene has no aliphatic carbons. They are all aromatic. No reference to naphthalene appears in the paper, I do not know what the newspaper writer was using as a source for this.
The explain what the research really did is that they created the conditions believed to be the source of much interstellar dust - outflows from carbon stars and measured its optical properties, and the type and proportions of bonds present. They did this because there is discrepancy between the amount of free carbon we can see in space, and the amount of carbon we believe should be there (for many reasons), and to clarify the dust absorption spectrum which scientists are still trying to fully understand.
They didn't discover any "new" types of carbon in space, what they did was improve the accounting for bond types, and produced a dust analog that closely matches unexplained absorption features. What the directly measured where the non-aromatic
carbon bonds, the ones that were aromatic are presumed with great certainty, given they had accounted for the other possibilities) to be those that were left.
No, they did not try to identify how large the carbon structures were, they were only counting bonds.
The summary and the article abstract do no make this clear but what that this device does is change the emissivity of the surface in the thermal IR band, which is around 10 microns. This means it simply alters the thermal emission - it makes is "white" or "black" (or shades of grey) in the IR band, which is far below normal optical bands, to match whatever the emission of the background is.
The optical band appearance of an object tells you nothing about its thermal IR band appearance. Two common materials that are pitch black in the thermal IR band: pure water and regular glass. Clean polished metal surfaces are reflective in both, but if it was could covered with an IR band black paint, and you would be none the wiser visually.
What effect would it have on body temperature regulation? Thermal radiation is only one of three ways your body disposes of metabolic heat under normal circumstances - the other two are convection and evaporation (which also works with convection but is a separate mechanism). The first two (radiation, convection) work both ways - you can gain heat, not lose it in a hot environment. There is also conduction but you have to be in physical contact with a surface for this to happen, so it is not the usual case for a camouflaged soldier unless wading (it is also in principle a two way process).
So modulating thermal emission changes only one of three (or four) ways of controlling core temperatures, requiring the others to take up the slack.The contribution of thermal emission to body temperature regulation varies with the environmental temperature. At about 33 C it is zero, since this is skin temperature. At higher environment temperatures you can only gain heat this way, not lose it - so being IR white is better. At low temperatures you can lose a lot, and if this bad being IR white is again helpful. But if you need to lose heat to keep from overheating, and the environment is below body temperature then being IR black is good. In general thermal radiation is the least important of the three (or four) processes, becoming dominant only when its cold and you will be snug in your parka anyway.
Is this is a new development, or a same old-same old story with a tweak?
This is looks like a genuine breakthrough of an amazingly cyberpunky kind, the sort of thing science fiction has been famous for "predicting" before it became available. The essential point is that thermal IR emissions are from the true surface of the material. You can't have a "layer of something" on top of the material with the variable emissivity, because then that would be the surface doing the emitting. You must genuinely change the surface properties with something behind the surface. This is a genuine IR optical "paint" that changes color electronically!
And the fact that it uses graphene, nanotubes, and ionic liquids combines buzz-word tech to the nth degree!
I read a lot of papers about the Fermi Paradox and most of them are worthless in addressing it - they typically take the form that "all civilizations will (or won't) do X and so the problem is solved" and in the "will" case is based on something that the only civilization known has not yet done. The assumption of some special behavior does nothing to address the paradox since it is universal in nature - it has to apply to all cases given that we see no evidence of civilizations at all.
This is one of the best papers I have seen, since it directly addresses the paradox, addresses the fundamental problem with analyzing it (uncertainty of the Drake parameters) and does not appeal to some special case solution. Also, given the valid formulation (many random variables with broad logarithmic distributions) their result isn't very sensitive to whether there are defects in the distribution modeling of any particular parameter.
But I do have a few of comments about their treatment, and about the subject as a whole.
First about the subject as a whole
Point one. Tthe probability of civilization in the Observable Universe is not relevant at all to the Fermi Paradox. Intelligent signals from distant galaxies would have to be broadcast to the entire Universe to have any useful chance of us being in the way of the signal to intercept it, which means the signal would have to have a power larger than actual stars. The Hubble telescope could not detect an optical signal at all as bright as the Sun at a billion light years, similar limits exist for radiotelescopes. Also, no intelligent civilization is going to be arriving here from distant galaxy groups. Even fusion propulsion will limit probes (that slow down) to about 5% c. The Universe has only been able to support long-term life for about the last 8 billion years (early Universe was too violent, heavy elements had to accumulate) so no probe from a civilization farther away than 400 million light years could even reach us. Thus even the most extreme volume relevant to the paradox is only something like 1/100,000 of the volume of the Universe, and I would argue really only the two large galaxies in the Local Group is relevant (the next closest group is 10 million light years away).
Point two. Arguments that we have "only one data point" when discussing life on Earth abuses the term "data point" horribly. We have only one system to observe, true, but it covers vast numbers of natural experiments across a billions of years, so it provides a great deal of data about the properties of living systems. In fact rare events in that history tell us a lot about the likelihood of those events.
Now about the paper. I have some objections to their modeling of two of their parameters, but the objections go in opposite directions and thus tend to cancel out to some extent.
Their distribution modeling of the origin of life is astounding. Due to our poor understanding of the processes of biogenesis (which they attempt to model) they assign a range of greater than 200 orders of magnitude (there are no more than 10^82 atoms in the Observable Universe). Given that life on Earth developed almost as soon as conditions permitted it (~100 million years after the end of sterilizing bombardments) actual evidence indicates that with suitable conditions is happens rapidly, and thus with high probability over a planet. It would be appropriate to consider the probability of a clone of Earth, but the approach they take to try to model biogenesis is not credible. In this case we do have actual evidence supporting a conclusion that it is likely with the right conditions. In other words, they are relying not so much on real uncertainties, but on poor arguments about uncertainty.
But by the same sort of consideration their uncertainty range for the probability of technological ("intelligent") life is implausibly large. They assign a log-uniform from 0.001 to 1 (based on "the literature", no discussion is offered), thus asserting that given life the likelih
Water is unique among small molecules. But there is some possibility I think that ammonia might be a suitable solvent for a different type of carbon-based life. They are both small polar molecules that are abundant, are often liquid over temperature ranges compatible with organics, and are both excellent solvents.
Slashdot Mirror
You don't get the cheapest anything without cutting corners.
It won't be able to accurately predict how biology will react to a new material
Except that the paper shows that it does. Not 100% accuracy, but 78-96%, quite good enough to do compound screening.
What sorts of "new material" do you think they will encounter? Chemicals made out of new elements?
Chemicals are put together is well defined ways, based on the small number of types of bonds that exist, though combinatorics leads to very large numbers of structures. Most organic molecules are made of just four types of atoms. A few others (phosphorus and sulfur, for example) are fairly common too. All chemicals share motifs, functional groups, and structural features with other chemicals. There are no "completely new" molecules.
(though it might be able to make a statistically better-than-chance guess).
It might indeed. In fact, it does better than standard non-chance prediction methods, and astronomically better than "chance" guesses. That is the whole point of the paper.
That's not machine learning. It's statistical analysis on a large scale.
And that is exactly what machine learning is. Discovering relationships in data patterns on very large sets of data with many attributes using statistical properties.
Molecules of interest generally have many parts ("functional groups") and many structural patterns. So many in fact that coming up with a consistent way of simply naming the molecules has been a challenge for chemistry for generations. And each of these groups and structural components interact with biological systems in different ways. After many decades of animal testing we have a very large set of data about the effects of millions of molecules on animals. This program finds structural pattern matches within this data that predict biological effects, and matches them to new molecules. If that isn't ML there is no ML.
It does not do away with animal testing. One needs to know the effects. Without animal testing you have no understanding of what the chemical will do. The only reason you would not need to perform animal testing is if testing has already been done and the new chemical has similar properties to previously tested chemicals.
It allows up use the vast database of animal tests that have already been done over a century or more. Every new chemical has "similar properties" to other chemicals, usually many chemicals - and many differences too. Finding how those similarities and differences relate to biological effects is what this program does.
Indeed. Melanin is not a fashion statement. It protects tissue from ultraviolet light.
The headline and article writer(s) for Phys.org is/are using the term "color" to mean pigment. The should have used "pigment" like the actual researchers do. The summary author here was just copying the Phys.org headline (mis)usage.
If they wanted to stay with "color" they should have used a compound term like "organic color", "biogenic color", or some-such.
All colors are real. Colors are the perceptions of the human visual system.
Appealing to "purple" as a special case of "color" is not well chosen since what is denoted "purple" is ambiguous (actually a problem with color names in general, when viewed cross-culturally). "Brown" is a much better example since it is necessarily and always a color mixture.
You are claiming I think that only spectral colors are real, which they are, but not exclusively. Color mixtures are still colors. Spectral colors are rarely seen under natural conditions (rainbows, and other occasions of color dispersion through refraction).
All codebases become inherited codebases if they remain in use.
I hope you get modded way up! This is the fallacy of the "but only use a subset and you are good".
Real world programmers must maintain and modify code bases that have been written by people that no one currently at the company even knows.
Things started to get better with the advent of the boost libraries and the C++11 standard brought a lot of the ideas from those libraries over in the standard.
Boost only came out about the start of 2000, at which point C++ had been under development by Stroustrup for 21 years, and the release of C++ 2.0, arguably the first "modern" version of C++ was 11 year earlier. And C++11 was only approved on 12 August 2011.
That is a lot pain and suffering and legacy code before the "better" arrived.
An odd lot, not my cup of tea, but I am scarcely in any position to judge. It is against-the-grain subcultures that help leaven the dough of a healthy society. And indeed the desire no to be surveilled is at the very heart of all liberty.
That is the refrain of an actual Insane Clown Posse song.
He3 makes things better, but I don't see how it helps enough.
Really it makes things far worse.
The only area where it helps is reducing the neutron damage and activation of the inner reactor parts, which are estimated to run only 5-8% of the capital contribution to the cost of electricity. But the reaction itself is ten thousand times harder to do (D/D fusion is only a few hundred times harder). We have good ideas for doing D/T fusion, that should work (at unaffordable cost) in several decades. We have none for He-3/D fusion, at this point it while the reaction is real, the technology is wishing for pink unicorns.
And then there is the fuel cost. One kilogram of D can be bought today for $3000. To get the equivalent amount of He-3 from the Moon you have to process ~300,000 tons of regolith on the Moon, and ship it back to Earth. Show me any sort of conceptual process that can do this for a penny a ton. Here on Earth currently total ore extraction and processing costs are in the range of $2 to $200 per ton, it is not going to be 100 to 10,000 times cheaper on the Moon, rather we can expect the reverse to be true.
Mapping the helium-3 distribution on the Moon is a worthy scientific endeavor - it will tell us much about how the solar wind interacts with the lunar surface.
But promoting the project for its "nuclear fuel" potential is so out of line with reality that it is deception, pure and simple.
First there is no prospect of building a helium-3 reactor. We currently cannot build a power-producing fusion reactor using the easiest fuel, deuterium-tritium, even though is reaction rate is ten thousand times faster than He-3/D at plausible temperatures.
Second we already can accurately forecast that when we can build a fusion reactor that uses that easiest to burn D-T fuel it will not be able to compete with any commercial source of electricity. The capital and operating costs of such a plant place the electricity cost at about ten times what wholesale electricity has been selling at for decades (an inflation adjusted current $30/MWh). This recent paper (accessible through Sci-hub) places the economics of a D-T plant in the best possible light and comes up with electricity costs due to the high capital cost of $175-$312 MWh*. Remember that He-3 fusion is ten thousand times harder, and we now have to mine the fuel on the Moon.
The only theoretical advantage of He-3 fusion is the lack of neutron emission from the main reaction (side reactions would still produce some). This would greatly reduce the neutron damage that requires periodic replacement of parts in D/T (or D/D) reactor, and greatly reduce the radioactive waste produced from neutron activated components. These are not major contributors to the projected cost of fusion power (the paper above assigns $14/MWh for these combined, 5-8% of the projected costs), so greatly reducing them does little to improve it.
And long before we can build a working He-3/D reactor, we will be able to build a D/D reactor using cheap, plentiful deuterium, available for a few thousand dollars a kilogram on Earth in effectively unlimited supply. The D/D reaction is "only" a few hundred times harder than D/T.
*The paper ultimately claims that it would be competitive, when externalities are costed, mostly by assigning very high externality costs to every other form of power, and assumes that all of that will be some day captured in electricity pricing. Its treatment of on-shore wind, and solar PV is especially suspect since it assigns levelized costs per MWh, 40 years in the future, that are several times higher than current, demonstrated costs now. This is a lot of special pleading.
This is the difference between investing and gambling.
...you make steel from ore, then you need coke...
This is currently true. There are no commercial processes to make iron without carbon as a reducing agent 2FeO + C -> 2Fe + CO2.
But iron can in principle be reduced electrolytically, like we do with aluminum.
There is no middle ground.
The process is currently under development, so there is no middle ground at the moment. This is the sort of thing TFA is discussing.
Nukes can work for cement, which just needs heat for the kiln. But nuclear aircraft? I don't think so. An iron blast furnace uses metallurgical coal (converted to coke), as an integral part of the process. You can't just drop in nuclear as a replacement.
The emission of CO2 from cement making is more fundamental to the process than it is for iron making.
Cement releases CO2 from the most fundamental chemical reaction required to make it: converting limestone to lime. This is the reaction: CaCO3 -> CaO + CO2. The release of carbon dioxide is unavoidable.
Iron on the other can in principle be made electrolytically although the process needs more development before it can be commercialized.
About half of the CO2 released gets recaptured by the cement over the course of several decades as the cement completes its curing process, but the other half never is. If we can capture CO2 at the cement plant source, that slow curing process might actually be a way to remove CO2 from the air.
Of course if we can capture cement plant CO2 at the source, we can do that with blast furnaces too, so that provides another option.
I am not a "sports fan" of any description, but to claim that "sports don't matter" is only valid if you also make the claim that "entertainment of any form doesn't matter".
I discovered a year or so back, using the iOS Google Maps for driving info, that they aren't actual maps - like you find on paper traffic maps - but very map-looking schematics that omit things they think you don't need to know about.
Like highway on-off ramps.
When you use the app for directions, where it knows your destination, yes, it shows you where to get off (if you follow its directions), but otherwise these traffic connections are not shown. I found this out since I wanted to see what the next off-ramp was, not having a specific destination programmed in. And I found none.
The Apple Maps app however did show every on-off ramp, like a real traffic map.
The true cosmically significant number to express ultimate enthusiasm is 42!!!!!!!!!!!!!!!!!!!!!!!!
And to anyone who is wondering, the enthusiasm limit imposed by /. is 24 (the number you see above). Try 25 and you get "filter error, try using fewer junk characters".
/. has standards! Who knew?
What the hell kind of question is that?!?!?
oh, dear
I'll bet you are expressing frustration at being denied the use of the Interrobang
I ended my sentence above with a string of Interrobangs, and they type in the editing window, but do not show in the preview. I will bet that /. is part of the anti-Interrobang conspiracy.
But not door-door delivery.
After driving through the Great Plains this past year, where small towns (and I mean small towns) are scattered over vast areas, and where there are many small towns that got so small they closed up shop, and hearing about the problems of getting goods delivered out there, I thought that a good strategy for a behemoth like Amazon that wants to Sell Everything to Everyone would be to set up a system of delivery hubs across the plains.
Buy up abandoned buildings in abandoned (or nearly so) towns, and convert them into the equivalent of post office boxes for packages, with locking storage. With web enabled smart locks the locals would not need dedicated containers, but would be notified which container had their stuff, and they would be able to unlock it to retrieve it.
An additional iteration would be along the lines of the TFA, allowing other locals to sign up for delivering to their neighbors.
Naphthalene has no aliphatic carbons. They are all aromatic. No reference to naphthalene appears in the paper, I do not know what the newspaper writer was using as a source for this.
The explain what the research really did is that they created the conditions believed to be the source of much interstellar dust - outflows from carbon stars and measured its optical properties, and the type and proportions of bonds present. They did this because there is discrepancy between the amount of free carbon we can see in space, and the amount of carbon we believe should be there (for many reasons), and to clarify the dust absorption spectrum which scientists are still trying to fully understand.
They didn't discover any "new" types of carbon in space, what they did was improve the accounting for bond types, and produced a dust analog that closely matches unexplained absorption features. What the directly measured where the non-aromatic
carbon bonds, the ones that were aromatic are presumed with great certainty, given they had accounted for the other possibilities) to be those that were left.
No, they did not try to identify how large the carbon structures were, they were only counting bonds.
The summary and the article abstract do no make this clear but what that this device does is change the emissivity of the surface in the thermal IR band, which is around 10 microns. This means it simply alters the thermal emission - it makes is "white" or "black" (or shades of grey) in the IR band, which is far below normal optical bands, to match whatever the emission of the background is.
The optical band appearance of an object tells you nothing about its thermal IR band appearance. Two common materials that are pitch black in the thermal IR band: pure water and regular glass. Clean polished metal surfaces are reflective in both, but if it was could covered with an IR band black paint, and you would be none the wiser visually.
What effect would it have on body temperature regulation? Thermal radiation is only one of three ways your body disposes of metabolic heat under normal circumstances - the other two are convection and evaporation (which also works with convection but is a separate mechanism). The first two (radiation, convection) work both ways - you can gain heat, not lose it in a hot environment. There is also conduction but you have to be in physical contact with a surface for this to happen, so it is not the usual case for a camouflaged soldier unless wading (it is also in principle a two way process).
So modulating thermal emission changes only one of three (or four) ways of controlling core temperatures, requiring the others to take up the slack.The contribution of thermal emission to body temperature regulation varies with the environmental temperature. At about 33 C it is zero, since this is skin temperature. At higher environment temperatures you can only gain heat this way, not lose it - so being IR white is better. At low temperatures you can lose a lot, and if this bad being IR white is again helpful. But if you need to lose heat to keep from overheating, and the environment is below body temperature then being IR black is good. In general thermal radiation is the least important of the three (or four) processes, becoming dominant only when its cold and you will be snug in your parka anyway.
Is this is a new development, or a same old-same old story with a tweak?
This is looks like a genuine breakthrough of an amazingly cyberpunky kind, the sort of thing science fiction has been famous for "predicting" before it became available. The essential point is that thermal IR emissions are from the true surface of the material. You can't have a "layer of something" on top of the material with the variable emissivity, because then that would be the surface doing the emitting. You must genuinely change the surface properties with something behind the surface. This is a genuine IR optical "paint" that changes color electronically!
And the fact that it uses graphene, nanotubes, and ionic liquids combines buzz-word tech to the nth degree!
You can get the article using Sci-hub.
I read a lot of papers about the Fermi Paradox and most of them are worthless in addressing it - they typically take the form that "all civilizations will (or won't) do X and so the problem is solved" and in the "will" case is based on something that the only civilization known has not yet done. The assumption of some special behavior does nothing to address the paradox since it is universal in nature - it has to apply to all cases given that we see no evidence of civilizations at all.
This is one of the best papers I have seen, since it directly addresses the paradox, addresses the fundamental problem with analyzing it (uncertainty of the Drake parameters) and does not appeal to some special case solution. Also, given the valid formulation (many random variables with broad logarithmic distributions) their result isn't very sensitive to whether there are defects in the distribution modeling of any particular parameter.
But I do have a few of comments about their treatment, and about the subject as a whole.
First about the subject as a whole
Point one. Tthe probability of civilization in the Observable Universe is not relevant at all to the Fermi Paradox. Intelligent signals from distant galaxies would have to be broadcast to the entire Universe to have any useful chance of us being in the way of the signal to intercept it, which means the signal would have to have a power larger than actual stars. The Hubble telescope could not detect an optical signal at all as bright as the Sun at a billion light years, similar limits exist for radiotelescopes. Also, no intelligent civilization is going to be arriving here from distant galaxy groups. Even fusion propulsion will limit probes (that slow down) to about 5% c. The Universe has only been able to support long-term life for about the last 8 billion years (early Universe was too violent, heavy elements had to accumulate) so no probe from a civilization farther away than 400 million light years could even reach us. Thus even the most extreme volume relevant to the paradox is only something like 1/100,000 of the volume of the Universe, and I would argue really only the two large galaxies in the Local Group is relevant (the next closest group is 10 million light years away).
Point two. Arguments that we have "only one data point" when discussing life on Earth abuses the term "data point" horribly. We have only one system to observe, true, but it covers vast numbers of natural experiments across a billions of years, so it provides a great deal of data about the properties of living systems. In fact rare events in that history tell us a lot about the likelihood of those events.
Now about the paper. I have some objections to their modeling of two of their parameters, but the objections go in opposite directions and thus tend to cancel out to some extent.
Their distribution modeling of the origin of life is astounding. Due to our poor understanding of the processes of biogenesis (which they attempt to model) they assign a range of greater than 200 orders of magnitude (there are no more than 10^82 atoms in the Observable Universe). Given that life on Earth developed almost as soon as conditions permitted it (~100 million years after the end of sterilizing bombardments) actual evidence indicates that with suitable conditions is happens rapidly, and thus with high probability over a planet. It would be appropriate to consider the probability of a clone of Earth, but the approach they take to try to model biogenesis is not credible. In this case we do have actual evidence supporting a conclusion that it is likely with the right conditions. In other words, they are relying not so much on real uncertainties, but on poor arguments about uncertainty.
But by the same sort of consideration their uncertainty range for the probability of technological ("intelligent") life is implausibly large. They assign a log-uniform from 0.001 to 1 (based on "the literature", no discussion is offered), thus asserting that given life the likelih
Water is unique among small molecules. But there is some possibility I think that ammonia might be a suitable solvent for a different type of carbon-based life. They are both small polar molecules that are abundant, are often liquid over temperature ranges compatible with organics, and are both excellent solvents.