3 ms·

> What similar opportunities are under our noses today? Material science is really hard, even though we already have at our disposal pretty much all the elemen

> What similar opportunities are under our noses today? Material science is really hard, even though we already have at our disposal pretty much all the elements we're likely ever going to have! There are so many combinations and conditions to try. I'd guess that if a room temperature and pressure superconductor is possible in our universe, it can likely be made in small amounts using today's technology. High-entropy alloys are another exciting space to explore. Materials science deserves a lot more respect than it gets. It's hard to appreciate how much hard-won materials advances, like neodymium magnets, enable the variety of technologies we have today.

kragen 4y ago
It might be more practical to change the temperature of the room! That's what they did for JWST. There might be more elements to draw on: antimatter, obviously, but also strange matter and metastable nuclear isomers. And although quantum dots aren't actually new elements, in many contexts they act like it. And don't assume that, just because the surface of a material is at atmospheric pressure, the material 10 microns below the surface is at a lower pressure than the core of the Earth. Really, though, as fascinating as materials science is, I think the biggest advances are elsewhere, at least for now and maybe for centuries. Better materials than 50 years ago weren't needed for Apollo 13, Linux, Stable Diffusion, seL4, Uber Cab, the WWW, AlphaZero, Z3, Hypothesis, Falcon, covid vaccines, or even 5G. (Better materials were used, but Linux displaying on a CRT with ten times bulkier chips would still be Linux. I mean, it was.)
- jbay808 4y ago
  Quantum dots, which can be manufactured with today's lab equipment from the elements on our periodic table, are a good example of why it's likely that the revolutionary materials of the future will also likely be something we can make with sufficiently clever application of today's tech. The stress in a pressure vessel goes as Pr/t, so for P=365 GPa and t= 10e-5m, the effective tensile strength of the surface would either have to be thousands of times higher than carbon fibre or else the total size of the superconducting particle would need to be microscopic. I'm not sure that a material with such stored energy wouldn't rearrange itself in short order to relieve that stress. Of course better materials than 50 years ago are not necessary for the Apollo missions, which took place 50 years ago. But there's also a materials reason why they didn't have quadcopter drones. We do still use most of the same structural materials today as back then, although materials science is one of the major limiting factors for rockets (especially reusable rockets) and metals that can survive high temperature oxidizing environments are a big step since then. And some of our most amazing materials, especially polymers, either were developed for or found their first applications in the Apollo program. For the internet, one of the greatest materials science revolutions in that period was the improvement in optical fibers, where losses dropped by something like 13 orders of magnitude. This was not initially even suspected to be possible. A good deal of semiconductor advances are arguably materials advances, and today silicon carbide power electronics are leading the way. Your smartphone has a flatscreen with a colour display thanks to the blue LED, another materials breakthrough of recent decades, which meant we could do away with bulky fluorescent backlights. It's not at all clear that we would have AlphaZero without the advanced materials driving today's semiconductor, magnetic, and optical materials, which enabled data storage and compute power to be cheaply available. Without this, it may have been only possible in the same sense that Da Vinci could design but not manufacture a helicopter in renaissance Italy.
  - kragen 4y ago
    Yup, all correct. Also LCDs themselves and in particular active-matrix LCDs are a recent materials advance. Also, for the kind of thick-walled microscopic pressure vessel I was describing, the situation is worse than sigma = Pr/t, or Pr/2t for spheres, which is the thin-wall approximation. Those structures are in the category of things the humans can be confident are possible, because the required pressure is only a few times higher than the tensile strength of buckytubes and nano-aggregated diamond, but have no idea how to make.
  - kragen 4y ago
    A couple of small quibbles about the optical fiber thing: 1. 17 dB/km attenuation fibers were first demonstrated 52 years ago, so better materials than precisely 50 years ago were not necessary for switching global telecom from copper over to fiber optics. That is, the materials science revolution in question wasn't "in that period", but immediately prior to it. Otherwise, everything you said is correct, as I said before. 2. Since the ARPANet started working in 01969, 53 years ago, long before the actual switchover to fiber optics, clearly wide-area packet switching wasn't dependent on fiber optics. When I first started using the internet in 01992, I was sharing a VAX with dozens of other students, all using a 64 kilobit per second frame-relay link, but the WWW was already up and running, and I used it at the time. Of course the NSFNet was running over optical fiber at the time, but if microwave waveguides had been the only available gigabit-capable communications medium, I'm confident they would have been laying microwave waveguides. Relay lines of microwave towers would also have worked. The crucial inventions of the WWW were HTML, HTTP, and the URL; from my point of view, those are fairly independent of computer speed and even bandwidth. I mean, yes, there's a minimal viable computer for parsing HTML; but as it turns out it's a Commodore PET running Contiki. All the improvements in computing speed and memory size in the last 45 years are unnecessary for running the WWW. You could imagine an optical-fiber-free internet where the server-centric nature of HTTP was an unacceptably heavy burden on the more limited backbone, especially with 6502-class CPUs constantly refetching pages in order to handle scrolling. Replacing HTTP with something like Usenet, BitTorrent, Freenet, Fastly, IPFS, or Kademlia is probably enough of a solution there. Maybe you wouldn't have YouTube or NetFlix, but that's why I didn't include YouTube or NetFlix in my list.

kragen 4y ago

It might be more practical to change the temperature of the room! That's what they did for JWST. There might be more elements to draw on: antimatter, obviously, but also strange matter and metastable nuclear isomers. And although quantum dots aren't actually new elements, in many contexts they act like it. And don't assume that, just because the surface of a material is at atmospheric pressure, the material 10 microns below the surface is at a lower pressure than the core of the Earth. Really, though, as fascinating as materials science is, I think the biggest advances are elsewhere, at least for now and maybe for centuries. Better materials than 50 years ago weren't needed for Apollo 13, Linux, Stable Diffusion, seL4, Uber Cab, the WWW, AlphaZero, Z3, Hypothesis, Falcon, covid vaccines, or even 5G. (Better materials were used, but Linux displaying on a CRT with ten times bulkier chips would still be Linux. I mean, it was.)

jbay808 4y ago
Quantum dots, which can be manufactured with today's lab equipment from the elements on our periodic table, are a good example of why it's likely that the revolutionary materials of the future will also likely be something we can make with sufficiently clever application of today's tech. The stress in a pressure vessel goes as Pr/t, so for P=365 GPa and t= 10e-5m, the effective tensile strength of the surface would either have to be thousands of times higher than carbon fibre or else the total size of the superconducting particle would need to be microscopic. I'm not sure that a material with such stored energy wouldn't rearrange itself in short order to relieve that stress. Of course better materials than 50 years ago are not necessary for the Apollo missions, which took place 50 years ago. But there's also a materials reason why they didn't have quadcopter drones. We do still use most of the same structural materials today as back then, although materials science is one of the major limiting factors for rockets (especially reusable rockets) and metals that can survive high temperature oxidizing environments are a big step since then. And some of our most amazing materials, especially polymers, either were developed for or found their first applications in the Apollo program. For the internet, one of the greatest materials science revolutions in that period was the improvement in optical fibers, where losses dropped by something like 13 orders of magnitude. This was not initially even suspected to be possible. A good deal of semiconductor advances are arguably materials advances, and today silicon carbide power electronics are leading the way. Your smartphone has a flatscreen with a colour display thanks to the blue LED, another materials breakthrough of recent decades, which meant we could do away with bulky fluorescent backlights. It's not at all clear that we would have AlphaZero without the advanced materials driving today's semiconductor, magnetic, and optical materials, which enabled data storage and compute power to be cheaply available. Without this, it may have been only possible in the same sense that Da Vinci could design but not manufacture a helicopter in renaissance Italy.
- kragen 4y ago
  Yup, all correct. Also LCDs themselves and in particular active-matrix LCDs are a recent materials advance. Also, for the kind of thick-walled microscopic pressure vessel I was describing, the situation is worse than sigma = Pr/t, or Pr/2t for spheres, which is the thin-wall approximation. Those structures are in the category of things the humans can be confident are possible, because the required pressure is only a few times higher than the tensile strength of buckytubes and nano-aggregated diamond, but have no idea how to make.
- kragen 4y ago
  A couple of small quibbles about the optical fiber thing: 1. 17 dB/km attenuation fibers were first demonstrated 52 years ago, so better materials than precisely 50 years ago were not necessary for switching global telecom from copper over to fiber optics. That is, the materials science revolution in question wasn't "in that period", but immediately prior to it. Otherwise, everything you said is correct, as I said before. 2. Since the ARPANet started working in 01969, 53 years ago, long before the actual switchover to fiber optics, clearly wide-area packet switching wasn't dependent on fiber optics. When I first started using the internet in 01992, I was sharing a VAX with dozens of other students, all using a 64 kilobit per second frame-relay link, but the WWW was already up and running, and I used it at the time. Of course the NSFNet was running over optical fiber at the time, but if microwave waveguides had been the only available gigabit-capable communications medium, I'm confident they would have been laying microwave waveguides. Relay lines of microwave towers would also have worked. The crucial inventions of the WWW were HTML, HTTP, and the URL; from my point of view, those are fairly independent of computer speed and even bandwidth. I mean, yes, there's a minimal viable computer for parsing HTML; but as it turns out it's a Commodore PET running Contiki. All the improvements in computing speed and memory size in the last 45 years are unnecessary for running the WWW. You could imagine an optical-fiber-free internet where the server-centric nature of HTTP was an unacceptably heavy burden on the more limited backbone, especially with 6502-class CPUs constantly refetching pages in order to handle scrolling. Replacing HTTP with something like Usenet, BitTorrent, Freenet, Fastly, IPFS, or Kademlia is probably enough of a solution there. Maybe you wouldn't have YouTube or NetFlix, but that's why I didn't include YouTube or NetFlix in my list.