I really like this post, I hope to see more like it on less wrong, and I strong-upvoted it.
Thanks, glad you liked it. You made quite the comment here, but I’ll try to respond to most of it.
Metal surfaces: Why not just build up the metal object in an oxygen-free environment, then add on an external passivation layer at then end?
To build up metal, you need to carry metal atoms somehow. That requires moving ions, because otherwise there’s no motive force for the transfer, plus your carrier would probably be stuck to the metal.
Without proteins carrying ions in water, this is difficult. The best version of what you’re proposing is probably directed electrochemical deposition in some solvent that has a wide electrochemical window and can dissolve some metal ions. Such solvents would denature proteins.
Inputs and outputs need to be transferred between compartments. Cells do use “airlock” type structures for transferring material, but some leakage would be inevitable.
The passivation layer could be engineered to be more stable than the naturally occurring oxidation layer the metal would normally have. There would still be a minimum size for metal objects, of course. (Or more precisely, a minimal curvature.) Corrosion could definitely be a problem, but Cathodic protection might help in some cases.
It’s true that proteins can be designed to bind strongly to metal oxide surfaces and inhibit corrosion fairly well. That’s actually an interesting research topic that might be useful for steel. But even that isn’t good enough on such a small scale, and you’d need to fully cover all exposed surfaces.
The only other option for “engineering” more-stable surfaces is metal nitrides or carbides, but that requires high temperatures, it’s not something enzymes can do.
Cathodic protection doesn’t help here. It doesn’t maintain a perfect equilibrium, and objects would still do Ostwald ripening and tend to become more spherical.
Agree that electrostatic motors are the way to go here. I’m not sure the power supply necessarily has to be an ion gradient, nor that the motor would otherwise need to be made from metal. Metal might be actively bad, because it allows the electrons to slosh around a lot.
I’m not sure what you mean by electrons “sloshing around”.
What about this general scheme for a motor?: Wheel-shaped molecules, with sites that can hold electrons. A high voltage coming from a power supply deposits electrons on one wheel, and produces holes on the other wheel. The corresponding sites are attracted to each other and once they get close enough, the electrons jump into the holes, filling them. Switching is determined by the proximity of various sites on the wheels as they move relative to each other. Considering how electrons are able to jump around between sites in the electron transport chain, this doesn’t seem impossible.
It’s certainly possible to make electromechanical computers with relays. And it’s possible to use MEMS electrostatic actuators for relays. They’re just not as good as semiconductors for computers. The MEMS relay approach is actually used in some devices for handling high-frequency radio signals.
Consider the analogous versions of ionic and electrostatic motors, and think about what’s better. Ionic motors use tubes filled with water instead of conductive wires with insulation; those transmit signals slower but are easier to make. Ionic motors can dump ions into solution instead of needing a conductor at a lower voltage. Ionic motors don’t have to deal with possible unintentional electrolysis. Ion gates are much easier to make with proteins than electrical switches.
Electrostatic motors are generally switched for each rotational step, but consider: If you want to compete with the energy usage of ionic motors, you can only use a few electron-volts per rotation. Semiconductor switches and relays are not so good for measuring out individual electrons.
Instead of sites that hold electrons, why not use sites that hold ions?
An intermediate design that I can imagine is a block with a series of tubes and chambers embedded in it. (The bulk of the block can hold electronics.) Most of the tubes are filled with water, so nanobot components can happily bounce around in them. But lots of components are also mounted to the walls of the tubes. You can’t clump together if you’re bonded to the wall of your tube. A small minority of tubes can be filled with gas, or even under vacuum, for any weird processes that may require those conditions. Pumping energy is volume times pressure, so the energy requirements could be reasonable as long as the volume is small.
That’s basically the same proposal as having gas-filled compartments in large cells, so this applies:
Any self-replicating cell must move material between outside and multiple compartments. Gas leakage by the transporters would be inevitable. Cellular vacuum pumps would require too much energy and may be impossible. Also, strongly binding the compounds used (eg CO2) to carriers at every step would require too much energy. (“Too much energy” means too much to be competitive with normal biological processes.)
If the objects bound to the walls of gas-filled compartments have movable arms, on a small enough scale, those arms would also get stuck to (or away from) the walls by electrostatic and dispersion forces.
The reason for doing things at high temperatures is to do reactions with a high activation energy. If we’re designing custom catalysts (artificial enzymes) for our nanobots, we can probably finesse it so that the enzyme coaxes the reactants into the high-energy intermediate state, even if the ambient temperature is low (via coupling to a more favorable reaction, for example).
Yes, enzymes can catalyze some difficult reactions. The main tools they use for that are hydrogen bonding patterns that stabilize specific conformations, and electrostatic fields around the active site. There are also p450 oxidases that put a reactive site in a hydrophobic pocket, and use it to oxidize hydrocarbons in a semi-controlled way to create a way to process them.
But enzymes aren’t magic. They have limitations. For example, methane can be metabolized by some bacteria, but it’s always oxidized to methanol in a reaction that consumes NAD(P)H. Half the energy of the methane is wasted to get it to a state that can be metabolized, and there’s just no way around that.
Another notable limitation of enzymes is the difficulty of making aliphatic hydrocarbons. That’s why hydrophobic stuff is almost always fatty acids or terpenes from DMAPP.
I’m fairly familiar with protein mechanisms and their limitations. Is there some other type of mechanism you’re proposing for low-temperature catalysis, something that enzymes don’t already use?
Also, covalent single-bonds can rotate, so there’s nothing preventing the existence of a covalently bonded structure that can also exhibit conformational changes.
Yes, proteins are covalently bonded. They’re also non-covalently bonded. If all their structure was covalent, then they wouldn’t be able to do conformational changes. And because some of it is non-covalent, they denature at high temperature.
Also, I’d guess that the stupidity of evolution has left a lot of low hanging fruit for humans. For example, rather than trying to do a reaction with proteins, we can do it with a group of complicated catalysts synthesized by proteins.
I’d bet on diamond synthesis still being possible somehow, but it does seem like a genuinely complicated question, so I’ll have to look into it further.
OK. Maybe you’ll learn something from the attempt.
Doesn’t have to be rigid, can still be connection based. For example, there could be simple protein-based building blocks that act like legos. An assembly head can assemble these, and move around on the surface of the part it’s building by accepting signals that correspond to “move 1 block left”, etc. Position is always exactly known, not because there’s a rigid beam anywhere in the system, but because we know the exact integer number of steps the assembly head has moved since the start.
OK, suppose you have a linear motor (like myosin) which is controlled by a signal (like a DNA sequence) that indicates a series of movements. (Something more computer-like would be less efficient than that). Also remember that on a molecular scale, energy-efficient = reversible. ATPase spins in both directions.
Compared to coding for a protein sequence, you’re using more information and more energy to do this. It’s also rather difficult to get single-protein-spacing-level control.
So, you’re imagining something like a protein with regularly spaced sites that can be attached to, and something that travels along it, with an enzyme-like tooltip that can bind to those sites to do a reaction that connects something. And that is...actually similar to how cytoskeletons work, but obviously they’re not directly controlled by DNA or RNA.
my general picture is that whenever a cell decides to make proteins somewhere and then transport them somewhere else, that costs genome space, which is very limited, so the cell can’t do that very often. Nanobots genuinely have different constraints from life here, in particular they have cheaper genome space, and so they can have custom designed pipes for every type of protein they use, and the pipe leads right to the chamber where that protein is being used. Huge information cost, but if it makes things work way better, it’s probably worthwhile for nanobots. I totally believe that life is using those techniques exactly as much as is optimal for it, though.
Specifying positions with positioners would require more bits of information than coding for proteins. DNA has high information density, and something much more compact than that wouldn’t have strong enough binding to be read accurately.
In what sense would nanobots have “cheaper genome space” than current cells? What mechanism do you envision being used for information storage?
I also don’t know why you’re scoffing at the potential for building computers here. The supposed “embodied” computation of existing cells is currently computing things that are keeping us alive, which is great, but you can’t exactly solve any other important problems on it. It’s not a flexible universal computer, in the sense of a Turing machine that can run any program.
If the point of your nanobots is to be “like current life, but worse, except it also produces a computer” then I think the usual word for that is “neurons”. The resulting computer would need to be better than current systems.
You want to grow brains that work more like CPUs. The computational paradigm used by CPUs is used because it’s conceptually easy to program, but it has some problems. Error tolerance is very poor; CPUs can be crashed by a single bit-flip from cosmic rays. CPUs also have less computational capacity than GPUs. Brains work more like...neural networks; perhaps that’s where the name came from.
Drexler envisioned such mechanical computers being used to control internal processes as well; that’s why I made the comparison. According to some people, this would be an advantage over how cells work for controlling internal operations, but I disagree.
Yep, I totally concede that the size and level of detail of exposed metal parts is going to be limited, the discussion would mostly be interesting in terms of whether or not nanomachines would be able to assemble large metal parts as an external product or metal parts that are fully embedded in another material (eg. copper wires embedded in diamond). The discussion about surface coatings and cathodic protection is just “haggling over the price”, so to speak.
I’m not sure what you mean by electrons “sloshing around”.
The thing where if you stick a piece of metal in an electric field, charges build up on the surface of the metal to oppose the field.
Consider the analogous versions of ionic and electrostatic motors, and think about what’s better. Ionic motors use tubes filled with water instead of conductive wires with insulation; those transmit signals slower but are easier to make. Ionic motors can dump ions into solution instead of needing a conductor at a lower voltage. Ionic motors don’t have to deal with possible unintentional electrolysis. Ion gates are much easier to make with proteins than electrical switches.
The original drawback I had in mind for ionic motors is that you need to drag a membrane everywhere that you want to use a motor, which is very inconvenient. Tubes are membranes, but they’re rolled up, which makes them a lot more convenient. Diffusion of ions is very fast on these scales, so I’d guess that ions and electrons are about equally good as power sources, unless the motor is going to be correspondingly very fast at using up lots of ions. On the other hand, I think you’re misunderstanding what I’m saying about the pure electrostatic motor. It doesn’t need any external switching electronics to power the motor, and in particular not silicon electronics. It should just spin given DC power of the correct voltage. The switching would work via proximity, and would happen on the wheel molecules themselves. It’s easiest for electrons to tunnel between two sites when they’re physically close in space. Depending on how the wheels are rotated relative to each other, various sites will be closer to or farther from each other in space, and this changes as the wheel spins.
If the objects bound to the walls of gas-filled compartments have movable arms, on a small enough scale, those arms would also get stuck to (or away from) the walls by electrostatic and dispersion forces.
Aren’t there lots of proteins that undergo conformational changes in ways that don’t look like “having arms”? Alternatively, I can make my arm negatively charged and put it on a tower-thingy made of lots of covalently bonded carbon so that it doesn’t bend. That tower-thingy holds it up away from the surface of the tube so it’s physically too far away to stick. But won’t it just stick to the tower-thingy? I’m one step ahead of you: I’ve also made the tower-thingy negatively charged!
I’m fairly familiar with protein mechanisms and their limitations. Is there some other type of mechanism you’re proposing for low-temperature catalysis, something that enzymes don’t already use?
I guess to start with, let’s say we’re making diamond. We’re building up a block of the stuff from carbon, and the dangling bonds on the edge of the structure connect to hydrogens. My first though would be a condensation reaction: We stick a methanediol onto the structure, replacing two dangling hydrogens with bonds to a new carbon atom. Two molecules of water are produced, made from the hydroxyl groups and those two hydrogens.
I think existing proteins can do condensation reactions okay, but maybe those become impossible when the carbon you’re trying to attach to is already bonded to three other carbons?
Yes, proteins are covalently bonded. They’re also non-covalently bonded. If all their structure was covalent, then they wouldn’t be able to do conformational changes. And because some of it is non-covalent, they denature at high temperature.
You’re correctly pointing out two extremes: very floppy chains with no additional structure, and fully rigid crystals with to flex at all. I’m trying to say that there exists an intermediate zone in between: Molecules that have enough covalent bonds to have structure, but there’s a few pivots where there’s just a single covalent bond that can rotate, and this gives the molecule flexibility to change shape. The structure of the molecule isn’t a chain, it’s a connected graph with lots of cycles.
I think the word for that is “cofactor”.
Yep, I know what a cofactor is. :) I’m saying that we could go farther in that direction than nature has. Normally a cofactor is held in a parent enzyme that still has to be just the right shape and everything. What if even the parent enzyme was just some organic molecule synthesized by other enzymes, rather than a protein itself? You complained above that any solvent that wouldn’t corrode metals would have to denature proteins. What if we could use molecules that aren’t proteins instead?
Also remember that on a molecular scale, energy-efficient = reversible. ATPase spins in both directions.
If we’re reading from RNA, near-reversibility seems fine and good. Maybe sometimes the assembly head takes a step backwards to where it came from and its RNA reader correspondingly takes a step backwards to the previous codon in the RNA sequence. If we’re working directly from electrical signals, maybe we have to just spend the energy needed to make sure that there’s no backtracking. When a tRNA detaches from its amino acid during protein synthesis, that seems pretty darn irreversible, but apparently the energy cost is bearable for the cell. If we’re now assembling larger lego bricks, each of which is an individual protein, probably the cost to make all the walking around on the surface irreversible is not too much compared to the cost to assemble each lego brick in the first place.
In what sense would nanobots have “cheaper genome space” than current cells? What mechanism do you envision being used for information storage?
A few senses. The first is that life is amazingly robust to errors, but not infinitely robust. The larger the genome, the more mutations you get per generation. Life does have some error repair mechanisms, but using information theory it’s possible to devise an error correcting code of arbitrarily high robustness. Life just can’t switch to using such a code because changing things up like that would mess with everything else. That kind of refactor is something that evolution just can’t do. If I recall correctly, evolutionary theorists do seem to think that this is a limiting factor on (coding) genome size.
Another sense is that nanobots can be much more cooperative rather than competitive. Each bacteria has to hold its entire genome, along with all the machinery needed to sustain and reproduce itself. Multicellular organisms have it much better in that their cells can cooperate and specialize, but even there, each cell has its own copy of the genome, even the genes it doesn’t particularly need at the moment. You could imagine that maybe only 1 in 1000 nanobots has to lug around the master copy of the genome, and the rest can just request the specific parts they need for their particular specialty they’re working on. They don’t out-compete the bot with the full genome because the system was designed top down and the presence of the the maser-genome holding bot makes sense from a top-down perspective. No need to worry about the nanobot equivalent of cancer because see above about error correction.
If the point of your nanobots is to be “like current life, but worse, except it also produces a computer” then I think the usual word for that is “neurons”. The resulting computer would need to be better than current systems.
Think about component size. Neurons are huge. Even current transistors are still much larger than individual proteins. The goal here is not “somehow build a computer with nanotechnology”. The goal is to build a computer whose logic gates are literally the size of molecules. (And also to make it reversible and therefore super power-efficient.) I know neurons are quite complex and are much more advanced than a simple transistor, but still, the size difference from neurons to molecules is ridiculous. And small components typically switch faster than large ones.
I guess to start with, let’s say we’re making diamond. We’re building up a block of the stuff from carbon, and the dangling bonds on the edge of the structure connect to hydrogens. My first though would be a condensation reaction: We stick a methanediol onto the structure, replacing two dangling hydrogens with bonds to a new carbon atom. Two molecules of water are produced, made from the hydroxyl groups and those two hydrogens.
I think existing proteins can do condensation reactions okay, but maybe those become impossible when the carbon you’re trying to attach to is already bonded to three other carbons?
Condensation reactions are only possible in certain circumstances. Maybe read about the mechanism of aldol condensation and get back to me. Also, methanediol is in equilibrium with formaldehyde in water.
I realize you don’t know my background, but if you want to say I’m wrong about something chemistry-related, you’ll have to put in a little more effort than that.
Thanks, glad you liked it. You made quite the comment here, but I’ll try to respond to most of it.
To build up metal, you need to carry metal atoms somehow. That requires moving ions, because otherwise there’s no motive force for the transfer, plus your carrier would probably be stuck to the metal.
Without proteins carrying ions in water, this is difficult. The best version of what you’re proposing is probably directed electrochemical deposition in some solvent that has a wide electrochemical window and can dissolve some metal ions. Such solvents would denature proteins.
Inputs and outputs need to be transferred between compartments. Cells do use “airlock” type structures for transferring material, but some leakage would be inevitable.
It’s true that proteins can be designed to bind strongly to metal oxide surfaces and inhibit corrosion fairly well. That’s actually an interesting research topic that might be useful for steel. But even that isn’t good enough on such a small scale, and you’d need to fully cover all exposed surfaces.
The only other option for “engineering” more-stable surfaces is metal nitrides or carbides, but that requires high temperatures, it’s not something enzymes can do.
Cathodic protection doesn’t help here. It doesn’t maintain a perfect equilibrium, and objects would still do Ostwald ripening and tend to become more spherical.
I’m not sure what you mean by electrons “sloshing around”.
It’s certainly possible to make electromechanical computers with relays. And it’s possible to use MEMS electrostatic actuators for relays. They’re just not as good as semiconductors for computers. The MEMS relay approach is actually used in some devices for handling high-frequency radio signals.
Consider the analogous versions of ionic and electrostatic motors, and think about what’s better. Ionic motors use tubes filled with water instead of conductive wires with insulation; those transmit signals slower but are easier to make. Ionic motors can dump ions into solution instead of needing a conductor at a lower voltage. Ionic motors don’t have to deal with possible unintentional electrolysis. Ion gates are much easier to make with proteins than electrical switches.
Electrostatic motors are generally switched for each rotational step, but consider: If you want to compete with the energy usage of ionic motors, you can only use a few electron-volts per rotation. Semiconductor switches and relays are not so good for measuring out individual electrons.
Instead of sites that hold electrons, why not use sites that hold ions?
That’s basically the same proposal as having gas-filled compartments in large cells, so this applies:
If the objects bound to the walls of gas-filled compartments have movable arms, on a small enough scale, those arms would also get stuck to (or away from) the walls by electrostatic and dispersion forces.
Yes, enzymes can catalyze some difficult reactions. The main tools they use for that are hydrogen bonding patterns that stabilize specific conformations, and electrostatic fields around the active site. There are also p450 oxidases that put a reactive site in a hydrophobic pocket, and use it to oxidize hydrocarbons in a semi-controlled way to create a way to process them.
But enzymes aren’t magic. They have limitations. For example, methane can be metabolized by some bacteria, but it’s always oxidized to methanol in a reaction that consumes NAD(P)H. Half the energy of the methane is wasted to get it to a state that can be metabolized, and there’s just no way around that.
Another notable limitation of enzymes is the difficulty of making aliphatic hydrocarbons. That’s why hydrophobic stuff is almost always fatty acids or terpenes from DMAPP.
I’m fairly familiar with protein mechanisms and their limitations. Is there some other type of mechanism you’re proposing for low-temperature catalysis, something that enzymes don’t already use?
Yes, proteins are covalently bonded. They’re also non-covalently bonded. If all their structure was covalent, then they wouldn’t be able to do conformational changes. And because some of it is non-covalent, they denature at high temperature.
I think the word for that is “cofactor”.
OK. Maybe you’ll learn something from the attempt.
OK, suppose you have a linear motor (like myosin) which is controlled by a signal (like a DNA sequence) that indicates a series of movements. (Something more computer-like would be less efficient than that). Also remember that on a molecular scale, energy-efficient = reversible. ATPase spins in both directions.
Compared to coding for a protein sequence, you’re using more information and more energy to do this. It’s also rather difficult to get single-protein-spacing-level control.
So, you’re imagining something like a protein with regularly spaced sites that can be attached to, and something that travels along it, with an enzyme-like tooltip that can bind to those sites to do a reaction that connects something. And that is...actually similar to how cytoskeletons work, but obviously they’re not directly controlled by DNA or RNA.
Specifying positions with positioners would require more bits of information than coding for proteins. DNA has high information density, and something much more compact than that wouldn’t have strong enough binding to be read accurately.
In what sense would nanobots have “cheaper genome space” than current cells? What mechanism do you envision being used for information storage?
If the point of your nanobots is to be “like current life, but worse, except it also produces a computer” then I think the usual word for that is “neurons”. The resulting computer would need to be better than current systems.
You want to grow brains that work more like CPUs. The computational paradigm used by CPUs is used because it’s conceptually easy to program, but it has some problems. Error tolerance is very poor; CPUs can be crashed by a single bit-flip from cosmic rays. CPUs also have less computational capacity than GPUs. Brains work more like...neural networks; perhaps that’s where the name came from.
Drexler envisioned such mechanical computers being used to control internal processes as well; that’s why I made the comparison. According to some people, this would be an advantage over how cells work for controlling internal operations, but I disagree.
Thanks for the detailed reply. Jumping right in:
Yep, I totally concede that the size and level of detail of exposed metal parts is going to be limited, the discussion would mostly be interesting in terms of whether or not nanomachines would be able to assemble large metal parts as an external product or metal parts that are fully embedded in another material (eg. copper wires embedded in diamond). The discussion about surface coatings and cathodic protection is just “haggling over the price”, so to speak.
The thing where if you stick a piece of metal in an electric field, charges build up on the surface of the metal to oppose the field.
The original drawback I had in mind for ionic motors is that you need to drag a membrane everywhere that you want to use a motor, which is very inconvenient. Tubes are membranes, but they’re rolled up, which makes them a lot more convenient. Diffusion of ions is very fast on these scales, so I’d guess that ions and electrons are about equally good as power sources, unless the motor is going to be correspondingly very fast at using up lots of ions. On the other hand, I think you’re misunderstanding what I’m saying about the pure electrostatic motor. It doesn’t need any external switching electronics to power the motor, and in particular not silicon electronics. It should just spin given DC power of the correct voltage. The switching would work via proximity, and would happen on the wheel molecules themselves. It’s easiest for electrons to tunnel between two sites when they’re physically close in space. Depending on how the wheels are rotated relative to each other, various sites will be closer to or farther from each other in space, and this changes as the wheel spins.
Aren’t there lots of proteins that undergo conformational changes in ways that don’t look like “having arms”? Alternatively, I can make my arm negatively charged and put it on a tower-thingy made of lots of covalently bonded carbon so that it doesn’t bend. That tower-thingy holds it up away from the surface of the tube so it’s physically too far away to stick. But won’t it just stick to the tower-thingy? I’m one step ahead of you: I’ve also made the tower-thingy negatively charged!
I guess to start with, let’s say we’re making diamond. We’re building up a block of the stuff from carbon, and the dangling bonds on the edge of the structure connect to hydrogens. My first though would be a condensation reaction: We stick a methanediol onto the structure, replacing two dangling hydrogens with bonds to a new carbon atom. Two molecules of water are produced, made from the hydroxyl groups and those two hydrogens.
I think existing proteins can do condensation reactions okay, but maybe those become impossible when the carbon you’re trying to attach to is already bonded to three other carbons?
You’re correctly pointing out two extremes: very floppy chains with no additional structure, and fully rigid crystals with to flex at all. I’m trying to say that there exists an intermediate zone in between: Molecules that have enough covalent bonds to have structure, but there’s a few pivots where there’s just a single covalent bond that can rotate, and this gives the molecule flexibility to change shape. The structure of the molecule isn’t a chain, it’s a connected graph with lots of cycles.
Yep, I know what a cofactor is. :) I’m saying that we could go farther in that direction than nature has. Normally a cofactor is held in a parent enzyme that still has to be just the right shape and everything. What if even the parent enzyme was just some organic molecule synthesized by other enzymes, rather than a protein itself? You complained above that any solvent that wouldn’t corrode metals would have to denature proteins. What if we could use molecules that aren’t proteins instead?
If we’re reading from RNA, near-reversibility seems fine and good. Maybe sometimes the assembly head takes a step backwards to where it came from and its RNA reader correspondingly takes a step backwards to the previous codon in the RNA sequence. If we’re working directly from electrical signals, maybe we have to just spend the energy needed to make sure that there’s no backtracking. When a tRNA detaches from its amino acid during protein synthesis, that seems pretty darn irreversible, but apparently the energy cost is bearable for the cell. If we’re now assembling larger lego bricks, each of which is an individual protein, probably the cost to make all the walking around on the surface irreversible is not too much compared to the cost to assemble each lego brick in the first place.
A few senses. The first is that life is amazingly robust to errors, but not infinitely robust. The larger the genome, the more mutations you get per generation. Life does have some error repair mechanisms, but using information theory it’s possible to devise an error correcting code of arbitrarily high robustness. Life just can’t switch to using such a code because changing things up like that would mess with everything else. That kind of refactor is something that evolution just can’t do. If I recall correctly, evolutionary theorists do seem to think that this is a limiting factor on (coding) genome size.
Another sense is that nanobots can be much more cooperative rather than competitive. Each bacteria has to hold its entire genome, along with all the machinery needed to sustain and reproduce itself. Multicellular organisms have it much better in that their cells can cooperate and specialize, but even there, each cell has its own copy of the genome, even the genes it doesn’t particularly need at the moment. You could imagine that maybe only 1 in 1000 nanobots has to lug around the master copy of the genome, and the rest can just request the specific parts they need for their particular specialty they’re working on. They don’t out-compete the bot with the full genome because the system was designed top down and the presence of the the maser-genome holding bot makes sense from a top-down perspective. No need to worry about the nanobot equivalent of cancer because see above about error correction.
Think about component size. Neurons are huge. Even current transistors are still much larger than individual proteins. The goal here is not “somehow build a computer with nanotechnology”. The goal is to build a computer whose logic gates are literally the size of molecules. (And also to make it reversible and therefore super power-efficient.) I know neurons are quite complex and are much more advanced than a simple transistor, but still, the size difference from neurons to molecules is ridiculous. And small components typically switch faster than large ones.
Condensation reactions are only possible in certain circumstances. Maybe read about the mechanism of aldol condensation and get back to me. Also, methanediol is in equilibrium with formaldehyde in water.
I realize you don’t know my background, but if you want to say I’m wrong about something chemistry-related, you’ll have to put in a little more effort than that.