Can someone explain nanotech enthusiasm to me? Like, I get that nanotech is one of the sci-fi technologies that’s actually physics-compliant, and furthermore it should be possible because biology.
But I get the impression that among transhumanist types slightly older than me, there’s a widespread expectation that it will lead to absolutely magical things on the scale of decades, and I don’t get where that comes from, even after picking up Engines of Creation.
I’m thinking of, e.g. Eliezer talking about how he wanted to design nanotechnology before he got into AI, or how he casually mentions nanotechnology as being one of the big ways a super-intelligent AI could take over the world. I always feel totally mystified when I come across something like that, like it’s a major gulf between me and slightly older nerds.
Trying for minimal technicalities:
There are at least 3 different technologies with not much surface-level usage similarities referred to as “nanotech”.
Assemblers: basically 3d printers, but way more flexible and able to make things like food, robots, or more assemblers.
Materials: diamondoids, buckytubes, circuitry. We already have some of these really, it’s just that we’d get more kinds of them, and they’d be really cheap to make with a nanotech assembler. Stronger, faster, more powerful versions of what modern tech already can do.
Nanobots, particularly medical: Basically can do all the things living cells can do, but better, and also being able to do most of the things machines can do, and commandable in exact detail. There are also a number of different ways they’d grant immortality enough that they are almost sure to do so even if most of them end up not working out.
Now you can ask questions about each one of these in order, with more specifics.
The question is about all these technologies—though it’s about 2 mainly insofar as 2 is an extension of 1.
So the question is why expect any of these technologies to mature on a timescale of decades?
(Or, assuming FOOM, why assume they’d be relatively low-hanging fruit for a FOOMing AI, such that “trick humans into building me nano assemblers” is a prime strategy for a boxed AI to escape?)
As I said, 2 is already here, and it’s becoming more here gradually.
For 3, we have a proof of concept to rip of: biological cells. Those also happens to have a specialized assembler in them already; the ribosome. And we can print instructions for it already. There’s only 1 problem left and that’s the protein folding problem. The protein folding problem is somewhat rapidly made progress on software wise, and even if that were to fail it won’t be all that long before we ca simply brute force it with computing power. Now, the other kinds of nanobots are less clear.
The assembler (1) is trickier; however, Drexler already sorta made a blueprint for one I think, and 3 will help a great deal with it as well.
For the fooming, it’s the 3 one, and ways to use it. As I said we already have the hardware, and things like the protein folding problem is exactly what an AI would be great at. Once it’s solved that it has full control over biology and can essentially make The Thing and/or a literal mind control virus, and take over that way.
There’s only 1 problem left and that’s the protein folding problem. The protein folding problem is somewhat rapidly made progress on software wise, and even if that were to fail it won’t be all that long before we ca simply brute force it with computing power.
Okay, so one sub-piece of puzzlement I have is why talk of protein folding as a problem that is either solved or unsolved—as if we (or more frighteningly, an AI) could suddenly go from barely being able to do it to 100% capable.
I was also under the impression that protein folding was mathematically horrible in a way that makes it unlikely to be brute forced any time soon, though I just now realized that I may have been thinking of the general problem of predicting chemistry from physics, maybe protein folding is much easier.
Predicting chemistry from physics should be easy with a quantum computer, but appears hard with a classical computer. Often people say that even once you make a classical approximation, ie, assume that the dynamics are easy on a classical computer, that the problem of finding the minimum energy state of a protein is NP-hard. That’s true, but a red herring, since the protein isn’t magically going to know the minimum energy state. Though it’s still possible that there’s some catalyst to push it into the right state, so simulating the dynamics in a vacuum won’t get you the right answer (cf prions). Anyhow, there’s some hope that evolution has found a good toolbox for designing proteins and that if can figure out the abstractions that evolution is using, it will all become easy. In particular, there are building blocks like the alpha helix. Certainly an engineer, whether evolution or us, doesn’t need to understand every protein, just know how to make enough.
I think the possibility that a sufficiently smart AI would quickly find an adequate toolbox for designing proteins is quite plausible. I don’t know what Eliezer means, but the possibility seems to me adequate for his arguments.
I’m not sure protein folding can be brute forced without quantum computers. There’s too many ways for it to fold. In real life, I’m pretty sure quantum tunneling gets involved. Simulations have worked, but I there’s a limit to that.
An analogy might help give a sense of scale here. This isn’t an argument, but it hints at the scope of the unknown unknowns in nanotech space. Here on our macroscopic scale, some wonders wrought by evolution include the smasher mantis shrimp’s kinetic attack, a bee hive’s eusocial organization, the peregrine falcon’s flight speed, and the eagle’s visual system. But evolution is literally mindless—by actually knowing how to do things, human engineering created electromagnetic railguns, networks of international trade, the SR-71 Blackbird, and the Hubble telescope. Now apply that kind of thinking to “because biology” on the nano scale...
Consider a machine as smart as a cellphone but the size of a blood cell.
In a sense, a protein or a drug is a smart molecule. It keys in to a very limited number of things and ignores the rest. There are many different smart proteins or smart molecules to be used as drugs with many different purposes. Even so, chemotherapy, for example, is primarily about ALMOST killing everything while differentially being a bit more toxic to the cancer cells.
Now increase the intelligence of ths smartest molecule by 10 fold, 100 fold, 1000 fold. Perhaps you give it the ability for a simple 2-way communication to the outside world. If its intelligence is increased, there should be MANY ways to allow it to distinguish a tumor, a micro-tumor, a cancerous cell, from all the good things in your body. All the sudden, the differential toxicity of “chemo” therapy (now nanotherapy) will be 10, 100X as high as it is for smart molecules.
Now consider these smart little machines doing surgery. Inoperable tumor? Not inoperable for a host of machines the size of bloodcells that will literally be able to operate on the most remote of tumors from inside of them.
Tendency towards obesity? How hard will it be to have a system of nanites that screw with your metabolism in such a way to eliminate all the stored fat in cells until told that, or until they measure that, we are down to a good level.
These are just a few stories from medicine. I expect anybody who does not wish to get sick and die would be enthusiastic about these, but YMMV.
The book had a gigantic impact on me. In a broad range of ways from hypertext through nanotech through various schemes for social organization and the long list of human needs such organization serves.
One of many things I liked was an illustration that economical improvement is not enough to make people live well. If I remember correctly, in the fictional world the food was free; anyone could go to a public matter-builder and take food from it. Yet some children were hungry… because their parents didn’t care enough for them to get outside of the house and bring them the food.
Moral of the story: however good situation you have, humans can make it bad by simply not caring even the smallest. (Unless we get to a situation where humans are replaced by robots completely.)
This seems to me like a hyperbole of the world we have now. Economically, the life in developed countries is so great that for most people who lived in the past it would be almost like a paradise. Yet we have a lot of suboptimality simply because people don’t care. (Maybe it’s because the wealth made social pressure less relevant, and many people just naturally don’t give a fuck about anything, and without the social pressure now they don’t even have to pretend.) The good life does not make us automatically stronger; it often makes us lazy. I believe the possibility is still there, but without outside pressure most people don’t care about becoming stronger.
Who, specifically are you talking about? I’m thinking of the extropians, who coined “transhumanism.” I’m not sure of the timeline; the original group was definitely into MNT before Stephenson, but maybe they expanded a lot after him, and maybe that was because of him.
Perhaps the reason is that the ideas we’re used to nowadays—like reconfiguring matter to make dirt and water into food or repair microcellular damage (for example, to selectively destroy cancer tumors) - were absolutely radical and totally unheard of when they were first proposed. As far as I know, Feynman was the first to seriously suggest that such a thing was possible, and most reactions to him at the time were basically either confusion, disbelief, or dismissal. Consider the average technologist in 1950. Hand-wound computer memories were state of the art, no one knew what DNA looked like, famines seemed a natural part of the order of things, and as far as everyone knew, the only major technological difference between the present and the future was maybe going to be space travel. Now someone comes along and tells you that there could be this new technology that allows you to store the library of congress in the head of a pin and carry out any chemical reaction just by writing down the formula—including the chemical reactions of life. The consequences would be, for instance, the ability to feed everyone on the planet at basically free. To you, such a technology would seem “Indistinguishable from magic.” Would it be a dramatic inferential step to then say that it could do stuff that literally is magic?
Nanotechnology never promised magic, of course. All it promises is the ability to rearrange atoms into a subset of those structures allowed by physics (a subset that is far larger than our current technology can do, but a subset nonetheless). It promised nothing more, nothing less. This is in itself dramatic enough, and it would allow all sorts of things that we probably couldn’t imagine today.
Can someone explain nanotech enthusiasm to me? Like, I get that nanotech is one of the sci-fi technologies that’s actually physics-compliant, and furthermore it should be possible because biology.
But I get the impression that among transhumanist types slightly older than me, there’s a widespread expectation that it will lead to absolutely magical things on the scale of decades, and I don’t get where that comes from, even after picking up Engines of Creation.
I’m thinking of, e.g. Eliezer talking about how he wanted to design nanotechnology before he got into AI, or how he casually mentions nanotechnology as being one of the big ways a super-intelligent AI could take over the world. I always feel totally mystified when I come across something like that, like it’s a major gulf between me and slightly older nerds.
Trying for minimal technicalities: There are at least 3 different technologies with not much surface-level usage similarities referred to as “nanotech”.
Assemblers: basically 3d printers, but way more flexible and able to make things like food, robots, or more assemblers.
Materials: diamondoids, buckytubes, circuitry. We already have some of these really, it’s just that we’d get more kinds of them, and they’d be really cheap to make with a nanotech assembler. Stronger, faster, more powerful versions of what modern tech already can do.
Nanobots, particularly medical: Basically can do all the things living cells can do, but better, and also being able to do most of the things machines can do, and commandable in exact detail. There are also a number of different ways they’d grant immortality enough that they are almost sure to do so even if most of them end up not working out.
Now you can ask questions about each one of these in order, with more specifics.
The question is about all these technologies—though it’s about 2 mainly insofar as 2 is an extension of 1.
So the question is why expect any of these technologies to mature on a timescale of decades?
(Or, assuming FOOM, why assume they’d be relatively low-hanging fruit for a FOOMing AI, such that “trick humans into building me nano assemblers” is a prime strategy for a boxed AI to escape?)
As I said, 2 is already here, and it’s becoming more here gradually.
For 3, we have a proof of concept to rip of: biological cells. Those also happens to have a specialized assembler in them already; the ribosome. And we can print instructions for it already. There’s only 1 problem left and that’s the protein folding problem. The protein folding problem is somewhat rapidly made progress on software wise, and even if that were to fail it won’t be all that long before we ca simply brute force it with computing power. Now, the other kinds of nanobots are less clear.
The assembler (1) is trickier; however, Drexler already sorta made a blueprint for one I think, and 3 will help a great deal with it as well.
For the fooming, it’s the 3 one, and ways to use it. As I said we already have the hardware, and things like the protein folding problem is exactly what an AI would be great at. Once it’s solved that it has full control over biology and can essentially make The Thing and/or a literal mind control virus, and take over that way.
Okay, so one sub-piece of puzzlement I have is why talk of protein folding as a problem that is either solved or unsolved—as if we (or more frighteningly, an AI) could suddenly go from barely being able to do it to 100% capable.
I was also under the impression that protein folding was mathematically horrible in a way that makes it unlikely to be brute forced any time soon, though I just now realized that I may have been thinking of the general problem of predicting chemistry from physics, maybe protein folding is much easier.
Predicting chemistry from physics should be easy with a quantum computer, but appears hard with a classical computer. Often people say that even once you make a classical approximation, ie, assume that the dynamics are easy on a classical computer, that the problem of finding the minimum energy state of a protein is NP-hard. That’s true, but a red herring, since the protein isn’t magically going to know the minimum energy state. Though it’s still possible that there’s some catalyst to push it into the right state, so simulating the dynamics in a vacuum won’t get you the right answer (cf prions). Anyhow, there’s some hope that evolution has found a good toolbox for designing proteins and that if can figure out the abstractions that evolution is using, it will all become easy. In particular, there are building blocks like the alpha helix. Certainly an engineer, whether evolution or us, doesn’t need to understand every protein, just know how to make enough.
I think the possibility that a sufficiently smart AI would quickly find an adequate toolbox for designing proteins is quite plausible. I don’t know what Eliezer means, but the possibility seems to me adequate for his arguments.
Ah, that’s helpful.
I’m not sure protein folding can be brute forced without quantum computers. There’s too many ways for it to fold. In real life, I’m pretty sure quantum tunneling gets involved. Simulations have worked, but I there’s a limit to that.
Try Nanosystems perhaps.
An analogy might help give a sense of scale here. This isn’t an argument, but it hints at the scope of the unknown unknowns in nanotech space. Here on our macroscopic scale, some wonders wrought by evolution include the smasher mantis shrimp’s kinetic attack, a bee hive’s eusocial organization, the peregrine falcon’s flight speed, and the eagle’s visual system. But evolution is literally mindless—by actually knowing how to do things, human engineering created electromagnetic railguns, networks of international trade, the SR-71 Blackbird, and the Hubble telescope. Now apply that kind of thinking to “because biology” on the nano scale...
Consider a machine as smart as a cellphone but the size of a blood cell.
In a sense, a protein or a drug is a smart molecule. It keys in to a very limited number of things and ignores the rest. There are many different smart proteins or smart molecules to be used as drugs with many different purposes. Even so, chemotherapy, for example, is primarily about ALMOST killing everything while differentially being a bit more toxic to the cancer cells.
Now increase the intelligence of ths smartest molecule by 10 fold, 100 fold, 1000 fold. Perhaps you give it the ability for a simple 2-way communication to the outside world. If its intelligence is increased, there should be MANY ways to allow it to distinguish a tumor, a micro-tumor, a cancerous cell, from all the good things in your body. All the sudden, the differential toxicity of “chemo” therapy (now nanotherapy) will be 10, 100X as high as it is for smart molecules.
Now consider these smart little machines doing surgery. Inoperable tumor? Not inoperable for a host of machines the size of bloodcells that will literally be able to operate on the most remote of tumors from inside of them.
Tendency towards obesity? How hard will it be to have a system of nanites that screw with your metabolism in such a way to eliminate all the stored fat in cells until told that, or until they measure that, we are down to a good level.
These are just a few stories from medicine. I expect anybody who does not wish to get sick and die would be enthusiastic about these, but YMMV.
Probably comes from Neal Stephenson’s The Diamond Age: Or, A Young Lady’s Illustrated Primer :-)
I definitely have found that this forum is NOT immune to fictional evidence.
I’m pretty sure that the people Chris is talking about are Stephenson’s source, not vice versa.
Eliezer definitely definitely seems to have caught the nano-enthusiasm bug pre-Diamond Age, but maybe the book had a big impact on other people?
The book had a gigantic impact on me. In a broad range of ways from hypertext through nanotech through various schemes for social organization and the long list of human needs such organization serves.
One of many things I liked was an illustration that economical improvement is not enough to make people live well. If I remember correctly, in the fictional world the food was free; anyone could go to a public matter-builder and take food from it. Yet some children were hungry… because their parents didn’t care enough for them to get outside of the house and bring them the food.
Moral of the story: however good situation you have, humans can make it bad by simply not caring even the smallest. (Unless we get to a situation where humans are replaced by robots completely.)
This seems to me like a hyperbole of the world we have now. Economically, the life in developed countries is so great that for most people who lived in the past it would be almost like a paradise. Yet we have a lot of suboptimality simply because people don’t care. (Maybe it’s because the wealth made social pressure less relevant, and many people just naturally don’t give a fuck about anything, and without the social pressure now they don’t even have to pretend.) The good life does not make us automatically stronger; it often makes us lazy. I believe the possibility is still there, but without outside pressure most people don’t care about becoming stronger.
Who, specifically are you talking about?
I’m thinking of the extropians, who coined “transhumanism.” I’m not sure of the timeline; the original group was definitely into MNT before Stephenson, but maybe they expanded a lot after him, and maybe that was because of him.
Perhaps the reason is that the ideas we’re used to nowadays—like reconfiguring matter to make dirt and water into food or repair microcellular damage (for example, to selectively destroy cancer tumors) - were absolutely radical and totally unheard of when they were first proposed. As far as I know, Feynman was the first to seriously suggest that such a thing was possible, and most reactions to him at the time were basically either confusion, disbelief, or dismissal. Consider the average technologist in 1950. Hand-wound computer memories were state of the art, no one knew what DNA looked like, famines seemed a natural part of the order of things, and as far as everyone knew, the only major technological difference between the present and the future was maybe going to be space travel. Now someone comes along and tells you that there could be this new technology that allows you to store the library of congress in the head of a pin and carry out any chemical reaction just by writing down the formula—including the chemical reactions of life. The consequences would be, for instance, the ability to feed everyone on the planet at basically free. To you, such a technology would seem “Indistinguishable from magic.” Would it be a dramatic inferential step to then say that it could do stuff that literally is magic?
Nanotechnology never promised magic, of course. All it promises is the ability to rearrange atoms into a subset of those structures allowed by physics (a subset that is far larger than our current technology can do, but a subset nonetheless). It promised nothing more, nothing less. This is in itself dramatic enough, and it would allow all sorts of things that we probably couldn’t imagine today.