Could a "Santa Claus Machine" ever be built?

[RuBisCO]

What do you think DNA does? It is actually a requirement to copy that less than perfect, otherwise a lot of evolution would be a much slower process. The error rate is a fairly constant number, just the how and what is sort of random.

Well I don't know if you want your products evolving.

[lajoswinkler]

It seems everyone here is forgetting one crucial thing. The number of particles in a chicken or a steak. Even if you could avoid Heisenberg principle (you can't) and even if you had enough energy for all those nuclear transmutations (because sheer inefficiency is unavoidable), it would take billions of years to synthesize even a cube of sugar, let alone something that complex and prone to decay as biological matter.

What do you think DNA does? It is actually a requirement to copy that less than perfect, otherwise a lot of evolution would be a much slower process. The error rate is a fairly constant number, just the how and what is sort of random.

DNA doesn't do anything. It's a bunch of molecular data. Protein machines are responsible for any DNA manipulation, and that is probably the most insanely complex molecular process in the known universe.

The whole system can't be compared to Santa Claus machine because it takes even more additional protein machines to make the final product - a functioning protein.

Santa Claus machine is a pipedream with infinite obstacles.

[3_bit]

I think with current tech fusing together your own atoms is just... unreasonable.

Yes, it could work, but the energy costs would be enormous, and the tech doesn't exist yet.

It would be cool, however, to turn dirt into diamonds, even though the real deal would likely be far cheaper.

[Seret]

It seems everyone here is forgetting one crucial thing. The number of particles in a chicken or a steak. Even if you could avoid Heisenberg principle (you can't) and even if you had enough energy for all those nuclear transmutations (because sheer inefficiency is unavoidable), it would take billions of years to synthesize even a cube of sugar, let alone something that complex and prone to decay as biological matter.

Depends on your method of assembly. Constructing macroscale objects from nanoscale components presupposes massive parallelism. That's not unrealistic, self-assembly techniques (or even commonplace chemistry) can coax molecules and structures to form in unimaginable numbers simultaneously. Sure, it's a huge leap from laying down a layer of thiols to building a beef steak, but given that we're only just finding our feet in this type of technology I wouldn't go so far as to rule it out.

[Camacha]

DNA doesn't do anything. It's a bunch of molecular data. Protein machines are responsible for any DNA manipulation, and that is probably the most insanely complex molecular process in the known universe.

In the literal sense you are correct, but you are also missing the point a bit.

Sure, it's a huge leap from laying down a layer of thiols to building a beef steak, but given that we're only just finding our feet in this type of technology I wouldn't go so far as to rule it out.

Not to mention our still pretty basic understanding - or even lack of understanding - of processes on the lowest levels. That makes assumptions so terribly dangerous. The best we can say is, that with our current understanding, it will be very hard to do.

Edited February 20, 2014 by Camacha

[ZetaX]

Those that claim that the uncertainty principle forbids any such thing would need to give a lot more of reason to go.

For one thing, it is not necessary to have perfect precission and generally, the allowed one is easily sufficient (your body doesn't place the position at subatomic exactness levels, so why would a replicator need to do¿). For another, the catch about the uncertainty principle is not just that you can't measure both properties with very high exactness at the same time, but that nothing "knows" them; this means that copying up to Planck exactness is as good as reality goes there. You still get such problems (look up some stuff about quantum information), but again those inadequacies are easily below any relevant scale.

So please argue this one fully if you think it is a hindrance (I think it isn't).

[Simon Ross]

Those that claim that the uncertainty principle forbids any such thing would need to give a lot more of reason to go.

For one thing, it is not necessary to have perfect precission and generally, the allowed one is easily sufficient (your body doesn't place the position at subatomic exactness levels, so why would a replicator need to do¿). For another, the catch about the uncertainty principle is not just that you can't measure both properties with very high exactness at the same time, but that nothing "knows" them; this means that copying up to Planck exactness is as good as reality goes there. You still get such problems (look up some stuff about quantum information), but again those inadequacies are easily below any relevant scale.

So please argue this one fully if you think it is a hindrance (I think it isn't).

Again, it comes down to the complexity of the object you wish to replicate as to the required accuracy of measurement needed.

It's why I used the house brick example. You will never be able to completely scan and replicate a house brick to 100% accuracy, but if you can get close enough it will to all intents and purposes still function as a house brick.

The problem really becomes relevant when you are trying to scan and replicate living structures as the complexity and required accuracy goes through the roof !

I'm usually loath to say anything is impossible, but I put replicating a human being with any accuracy out there with time travel and gravity control

[lajoswinkler]

Depends on your method of assembly. Constructing macroscale objects from nanoscale components presupposes massive parallelism. That's not unrealistic, self-assembly techniques (or even commonplace chemistry) can coax molecules and structures to form in unimaginable numbers simultaneously. Sure, it's a huge leap from laying down a layer of thiols to building a beef steak, but given that we're only just finding our feet in this type of technology I wouldn't go so far as to rule it out.

Ionic crystals, metallic lattices, that might work by pure repetition (although it's also insanely difficult). Biological matter? Never.

Those that claim that the uncertainty principle forbids any such thing would need to give a lot more of reason to go.

For one thing, it is not necessary to have perfect precission and generally, the allowed one is easily sufficient (your body doesn't place the position at subatomic exactness levels, so why would a replicator need to do¿). For another, the catch about the uncertainty principle is not just that you can't measure both properties with very high exactness at the same time, but that nothing "knows" them; this means that copying up to Planck exactness is as good as reality goes there. You still get such problems (look up some stuff about quantum information), but again those inadequacies are easily below any relevant scale.

So please argue this one fully if you think it is a hindrance (I think it isn't).

Those inadequacies are the difference between a pile of smelly goo and a bunch of cells, even dead ones. Even if you could be precise enough, you'd need to do it instantly. Position, conformation and energy of protein molecules are everything for biological matter.

If we wanted to replicate a raw steak, we're talking about number of atoms on the order of magnitude of 10²³-10²⁴, and it all (positions, energies) needs to be done instantly. It's impossible not only by the means of technology (what computer could store all that information? It would have to be big as a planet). It's scientifically impossible.

Edited February 20, 2014 by lajoswinkler

[ZetaX]

Please read my post again and/or check the numbers: I was never and nowhere talking about complexity and I don't care about it there; claim was that QM makes it impossible, and my counterclaim is that this is wrong. Other reasons are irrelevant for that (but the complexity one is also possibly wrong, but for more subtle reasons).

The exactness allowed by QM is completely inside the exactness needed to recreate proteins. It has to be even without checking numbers as otherwise proteins would magically break the uncertainty principle to require more exactness than we can ever hope to measure.

[Seret]

Ionic crystals, metallic lattices, that might work by pure repetition (although it's also insanely difficult). Biological matter? Never.

Never? Biological systems manage to build more biological systems by constructing them from scratch. It's completely doable, even if it's not something we're anywhere near capable of doing ourselves.

needs to be done instantly

No, it just needs to either be kept alive or for decay to be stalled during construction. At the most that adds overheads, but it doesn't rule it out. Bear in mind that it's normal to age good beef for a few weeks anyway, so taking that long wouldn't be an issue

Edited February 21, 2014 by Seret

[lajoswinkler]

Please read my post again and/or check the numbers: I was never and nowhere talking about complexity and I don't care about it there; claim was that QM makes it impossible, and my counterclaim is that this is wrong. Other reasons are irrelevant for that (but the complexity one is also possibly wrong, but for more subtle reasons).

The exactness allowed by QM is completely inside the exactness needed to recreate proteins. It has to be even without checking numbers as otherwise proteins would magically break the uncertainty principle to require more exactness than we can ever hope to measure.

If we are talking about going "from the scratch" i.e. going from subatomic particles, then uncertainty principle plays a huge role. You can't deny that.

You also can't deny the huge numbers involved. It's easily calculated.

Never? Biological systems manage to build more biological systems by constructing them from scratch. It's completely doable, even if it's not something we're anywhere near capable of doing ourselves.

We aren't talking about the stuff that's routinely done in genetic laboratories in those small tubes. We're talking about subatomic particle manipulation to make tissues. Tissues are made from cells, and this is a grossly oversimplified surface of a cell.

That will never be possible to replicate directly. The number of atoms is unbelieveable, and more important, you'd need a computer to calculate the energies and conformations of each protein molecule.

No, it just needs to either be kept alive or for decay to be stalled during construction. At the most that adds overheads, but it doesn't rule it out. Bear in mind that it's normal to age good beef for a few weeks anyway, so taking that long wouldn't be an issue

I don't think you realize how complex this thing is.

Just how are you going to keep it from decay while not precipitating the proteins inside? "Stasis chamber"? "Decay compensator ray"? LOL

I really think many of you here have a narrow, mechanical engineering approach to these things. This is not a matter of technological advancement.

At best we can hope for some kind of matter lasers that could build microscopic amounts of simple lattices, using prepared atoms.

[Seret]

We aren't talking about the stuff that's routinely done in genetic laboratories in those small tubes. We're talking about subatomic particle manipulation to make tissues. Tissues are made from cells, and this is a grossly oversimplified surface of a cell.

I don't think subatomic manipulation is either possible or desirable to achieve the OP's objective of a "build anything" machine. You're going to need to building with molecules, or at the most atoms. That means you're working in the realms of biology, nanotech or chemistry, not witchcraft.

Just how are you going to keep it from decay while not precipitating the proteins inside? "Stasis chamber"? "Decay compensator ray"? LOL

Refrigeration would do, some nanoscale assembly processes actually work better at lower temperatures, since thermal noise is a bit of an obstacle.

[Camacha]

That will never be possible to replicate directly. The number of atoms is unbelieveable, and more important, you'd need a computer to calculate the energies and conformations of each protein molecule.

We were going to need a computer anyway, so I do not see the point. You just state that is will never be possible, but I do not see any reasons why - other than that it is a lot of work and matters are very complex. Sure. Both those things are true. Neither rules it out.

I really think many of you here have a narrow, mechanical engineering approach to these things. This is not a matter of technological advancement.

I think exactly the same for the first sentence, and exactly the opposite for the second. I think its quite dubious to call things impossible when our best knowledge at the moment is, very generously put, lacking when it comes to the domain of very small things. At the same time, a solution to this problem will probably be highly complex and a lot more advanced than the aforementioned atomic brick layer.

Doesn't the domain of very small things sound like something from a movie or Monty Python sketch?

[lajoswinkler]

Guys, do you know what is a mole (unit)?

[Camacha]

Guys, do you know what is a mole (unit)?

It is a SI base unit.

[Sillychris]

Yes, it would take enormous energy and yes, it would take technology we haven't thought of, but there is nothing in physics that says you can't rearrange nuclear matter at will.

Ultimately, it's an engineering problem. Your Santa Claus machine is possible, except...

You are limited by quantum mechanics and Heisenberg's uncertainty principle. It would be impossible to obtain an exact particle by particle code to replicate an object because the details of individual particles are fuzzy and do not exist as exact values. The reason things seem so concrete to us at a macroscopic level is because the quantum fuzziness gets ironed out by Avogadro's number.

That being said, I am convinced you could get pretty close. Bucket of chicken from air? It doesn't need to be exact. I'm ok if a few thousand protons are in the wrong place on my drumstick. Besides, I wouldn't be able to tell. I'm macroscopic!

[lajoswinkler]

It is a SI base unit.

It was kind of a rhetorical question.

Yes, it would take enormous energy and yes, it would take technology we haven't thought of, but there is nothing in physics that says you can't rearrange nuclear matter at will.

Ultimately, it's an engineering problem. Your Santa Claus machine is possible, except...

You are limited by quantum mechanics and Heisenberg's uncertainty principle. It would be impossible to obtain an exact particle by particle code to replicate an object because the details of individual particles are fuzzy and do not exist as exact values. The reason things seem so concrete to us at a macroscopic level is because the quantum fuzziness gets ironed out by Avogadro's number.

That being said, I am convinced you could get pretty close. Bucket of chicken from air? It doesn't need to be exact. I'm ok if a few thousand protons are in the wrong place on my drumstick. Besides, I wouldn't be able to tell. I'm macroscopic!

If you're limited by Heisenberg, then it's not an engineering problem, is it?

You're wrong. You could tell the difference because if something goes wrong on quantum levels, macroscopic properties change dramatically.

[Simon Ross]

Yes, it would take enormous energy and yes, it would take technology we haven't thought of, but there is nothing in physics that says you can't rearrange nuclear matter at will.

Ultimately, it's an engineering problem. Your Santa Claus machine is possible, except...

You are limited by quantum mechanics and Heisenberg's uncertainty principle. It would be impossible to obtain an exact particle by particle code to replicate an object because the details of individual particles are fuzzy and do not exist as exact values. The reason things seem so concrete to us at a macroscopic level is because the quantum fuzziness gets ironed out by Avogadro's number.

That being said, I am convinced you could get pretty close. Bucket of chicken from air? It doesn't need to be exact. I'm ok if a few thousand protons are in the wrong place on my drumstick. Besides, I wouldn't be able to tell. I'm macroscopic!

Unfortunately if you don't get the base information correct at the molecular level, the errors propagate and exponentially multiply into the macroscopic level. It's very straightforward math.

Your not going to end up with a bucket of chicken, your not even going to end up with something resembling a single chicken cell

[Seret]

I think he was talking about there being a small but tolerable error rate in the process, not a fundamental error that affected every molecule. A few thousand atoms or molecules that didn't come out right would indeed be no problem. You don't get 100% perfect molecular assembly in chemistry or biology either.

[Simon Ross]

I think he was talking about there being a small but tolerable error rate in the process, not a fundamental error that affected every molecule. A few thousand atoms or molecules that didn't come out right would indeed be no problem. You don't get 100% perfect molecular assembly in chemistry or biology either.

OK, so lets just break this down to what needs to be actually done to produce say a chicken wing...

You have to define it at the macro level, you then need to define it at the cellular level, you then need to define it at the molecule level, you then need to define it at the atomic level

Not saying it is impossible, but trust me, it's a LONG way from where we are and will conceivably be any time soon

[nhnifong]

There is a program that can make anything in conway's game of life out of gliders. It places them all over the place and then you run the simulation and they all crash into eachother in a carefully choreographed way and produce the desired pattern. Perhaps there is a particle in the standard model that could do something similar.

[Sillychris]

OK, so lets just break this down to what needs to be actually done to produce say a chicken wing...

You have to define it at the macro level, you then need to define it at the cellular level, you then need to define it at the molecule level, you then need to define it at the atomic level

Not saying it is impossible, but trust me, it's a LONG way from where we are and will conceivably be any time soon

The question was whether it's possible, not whether we're close.

[Simon Ross]

The question was whether it's possible, not whether we're close.

Well please tell us how we do it ?

[Sillychris]

Well please tell us how we do it ?

All ordinary matter is made out of protons, neutrons and electrons (actually up quarks, down quarks, and electrons but that's deeper than we need for transmutation purposes).

Over the course of the 20th century, man had become increasingly more adept at transmutation. Gold from lead is now possible, as it is merely a rearrangement of the nuclear matter.

In fact, high purity gold IS manufactured by neutron bombardment of mercury for extremely high purity scientific applications.

Back to your question: How do we do it? By putting the protons and neutrons where the recipe says they need to be. Electrons are plentiful and will just show up on their own.

How does one arrange protons and neutrons? With forces. There are four forces we know of... well, actually 3 but for all practical purposes at our energy levels there are 4: Nuclear strong force, Nuclear weak force, electromagnetic force and gravity. Master manipulation of the four forces and build your bucket of chicken out of air.

You want more specific? This is distant future technology. How am I supposed to guess what kind of technology could exist in millions of years?

However, physicists have become quite proficient at manipulating individual atoms in recent years. They have been quite proficient at changing atoms from one element to another since pre-ww2. Those are the two skills required for a Santa-Claus machine. Maybe we aren't as far as I would wager...

[Seret]

Maybe we aren't as far as I would wager...

The thing is, in reality the required technologies to make this happen wouldn't be developed, unless it was for another application. It's never going to be cheaper to fabricate macroscale objects from subatomic components than it is to use a process operating at a higher level.

Assembly at molecular and atomic scales is feasible, we do this now to a greater or lesser degree. But there's no real incentive to build at a subatomic level unless you require absurd levels of precision, which just isn't necessary to create a macroscale object. Manipulation at a subatomic level requires manipulating strong forces with high levels of control, which is always going to be more expensive. Such manipulation might be useful as a research tool, but it's won't be economical as a production process for macroscale objects.

So even if it were physically possible to build such a machine, it's not certain that anybody would actually want to. I personally think the idea of working at a subatomic level shifts the whole concept into the realm of sci-fi. Assembly at the atomic level however is a lot more believable, since that's actually the goal of current research.

Edited February 23, 2014 by Seret