Monday, August 17, 2009

Tough Guide: Nanotech

There are two types of nano-scaled machines in the Known Galaxy. One kind does the same things that other machines do, only smaller. They scrape scale off the insides of microtubes, or the plaque off your teeth, which is why dental assistants are unknown on the more advanced Planets. Nano-touchscreens were a failure, because even Really Aliens lack tentacles small enough to use them. Taken as a group, such nanomachines are pervasive and commonplace, fascinating to tech geeks but ignored by everyone else, especially since you can't see them except through an expensive nanoscope.

The other kind of nanomachine does everything, and does it faster and cheaper, leaving the mere laws of physics and mechanics in the dust - well, not quite dust. Especially, these nanomachines make more nanomachines, which ought to give any Planets that develop this form of nanotech an enormous industrial advantage over everyone else. Unfortunately, what nanomachines of this type do best is to turn everything in reach into gray goo, including Planets so careless as to invent them. This must effectively limit their spread, since the Known Galaxy has not yet been turned into gray goo.

Nanotech of the gray goo type seems to have had a brief period of vogue in SF in the years around the turn of the third millenium CE. After a short time, however, the nanotech subgenre vanished, leaving no identifiable traces. It is speculated that all such works, along with their authors, were turned into gray goo.


Related links: Here is the original Tough Guide to the Known Galaxy. This blog previously took on Elevators and the Singularity.

34 comments:

Anonymous said...

Nanotechnology has been one of those "magical" technologies that were seen as an all powerful answer that can seemingly do the impossible without much effort.

If only because much of what nanotechnology claims to do really is impossible. As outlined and explained in detail on the following web page:

http://www.stardestroyer.net/Empire/Tech/Myths/Nanotech.html

The only true logical usage of nanotechnology such as its mainstay of nanomachines are through medical and computing use. Everything else is simply too grand a scale for the passe tech to perform and would require a sense of "magic" to perform. And no, I don't mean the quote about "Any sufficiently advanced technology can be interpreted as magic", I mean Dungeons and Dragon, Lord of the Rings kind of magic.

- Sabersonic

Anonymous said...

Nanotech does seem to have retreated to backstage these days. It's refered to but rarely seen, like its bacterial kin.

I think SF misses out on a lot by ignoring micro-robots. Really BIG machines are a long-time favourite of geeks, and really TINY machines are still reasonably popular if not as much as they used to be. But SF seems to ignore the bug-sized (Or even cat/small dog sized) robots, which is unfortunate because they seem to have some very useful applications and also because a swarm of ratbots armed with cutting tools and sharp mandibles would be really cool.

Ian_M

Jean Remy said...

Interestingly enough I was listening to NPR on Sunday. The show "On the Media" had an interview with Ray Kurzwell. Of course he talked about his little baby, the Singularity, and was not shy about throwing nanomachines in the mix.

What strikes me is that the entire basis for his future predictions are based solely on statistical data of relative progress over the last 50 years or so, and even then seemingly only in the matter of information technology. What strikes me is that, for a futurist, he is far too much in love with mathematical abstract and rather out of touch with fields of history, sociology and psychology. He perhaps even misses a few steps in biology or even physics.

"What used to fit in a building now fits in your pocket, and what fits in your pocket today will fit inside a blood cell in 25 years" he says. But cellphones have been fitting in our pockets for quite some time, and though they now have more bells and whistles, there's still a size at which the device become impractical. We're not going to have phones the size of our thumbnails because, to be frank, I already lose my phone enough times a day. I need it to be a certain size, so the model breaks down. If it was possible to fit a cellphone in a blood cell (how do you power it and what to use as an antenna and what exactly is the strict minimum size of such a device as limited by physics or even quantum mechanics?) the fact is there will have to be a break in his statistical curve. That is, he purports that there is a predictable miniaturization curve for the device, but if there are no devices below a certain size or above another, then there clearly are factors that do not fit within the simplistic mathematical expression.

Here, the factor is human: no one will buy a phone the size of a thumbnail, especially not people who have ever used contacts and know how easy it would be to lose one of these.

I would postulate the same for nanobots, that psychological and sociological factors would undermine the perfect mathematical model because, well, people are not machines and do not really fit tidy statistical curves (ask any polling agency around election time. Oh, they try. They try. They fail, though) In fact I would argue that the more oddball the technology is, the greater the resistance to it. A cellphone is a phone, and phones have been around since Bell, so it is something we are reasonably comfortable with, but an invisible robot in your brain is "spooky", and 20 years is far to little time to get used to the idea.

Especially when the first nanobots turn out to have really buggy programming. My computer crashes enough as it is, do we really want to crash our brains? And in that case, will the Blue Screen of Death jokes cease to be funny? (or become more funny in a very sick way?)

Anonymous said...

@Jean Remy

Yeah, the problem with singularitarianism is that it ignores that most techs tend to undergo exponential development for a while, and then plateau. The statistical trend won't continue forever.

I think you misunderstand Kurzwell's statement about "What fits in your pocket now will fit inside a blood cell in 25 years." He's talking about raw computing power, not about the function of the device. What he was talking about is the fact that a cellphone now has the processing power of the mainframes of 60 years ago, and (he believes), someday you will be able to fit that sort of processing power into the size of a blood cell. We will (assuming his predictions pan out) still have cell phones, at about the same size, but they will have the processing power that a data center full of servers has now (although, given the classical singularitarian assumptions, we might not be the kind of beings that would *need* cell phones by that point).

Linguofreak

Jean Remy said...

I took his statement to mean: "we'll have cellphone the size of a blood cell in our brains." but you might be right in that he meant "raw computing power". However I want to point out that even though we can fit more computing power in a smaller package, my computer hasn't really gotten smaller since my old 8086. That's because of several factors:

1/ The most obvious: my computer is a lot more powerful for the same volume of space. So while we might get to a point where we can fit my current 3MHz chip on a blood cell, I'm sure my future computer will still be quite big, just a LOT more powerful. In other words: miniaturization is more about fitting more stuff in the same volume than actually reducing the overall volume.

2/ The more subtle: The power source for my computer is still the exact same clunky piece of machinery it was 20 years ago. The heat sinks and fans are equally as big also. I say it is more subtle because it's not the first thing we think about when we talk about a computer. But there is a minimum practical size for a transformer and attendant heat sinks, fans, power cables. You simply can't miniaturize those (very well). In other words: our computing power is miniaturized, but still depends on older technologies that either have not been, or cannot be miniaturized, because of physical constraints. A heat sink has to be a certain size and coupled with a fan of a certain size to evacuate a certain amount of heat over a certain time, based on fundamental physical laws.

3/ The last but not the least: intercommunication. Inside a computer wires and printed circuits, outside either a cable modem or a wireless one with an antenna. What about between nanobots? Of course wires and circuits are out. The problem is that wireless communications require an antenna of a size reflecting the frequency you are using. Keeping in mind that perhaps using microwave radiation inside your body is perhaps not the greatest idea ever. And that your body doesn't make that great of an antenna itself.

So while we could potentially pile transistors into a blood cell, that doesn't necessarily mean that nanomachines are necessarily viable in terms of every day "internet in your brain" use which Ray Kurzweil sees as the ultimate goal of nanotech.

BTW full transcript of that interview is here:
http://www.onthemedia.org
/transcripts/2009/08/14/03

Anonymous said...

A couple remarks:

There is a bit of a trend towards computers getting smaller. Your desktop computer of today may not be much smaller than your old 8086, but there are still laptops and cell phones that are quite small and pack more computing power than that 8086. When the 8086 came out, there were no Laptops or cell phones. Laptops are especially significant here, since they fill roughly the same niche as desktops, but are smaller and more portable.

With regards to your remark about cooling and power supply, bear in mind that a cell phone boasts more computing power than your old 8086, but has only a tiny battery for power, and no fans, only passive cooling. The amount of power and cooling needed for a given amount of processing power has gone down. Really, point number 2 is more a consequence of point number 1, rather than a separate arguement. Your desktop today has roughly equivalent power and cooling requirements to your old 8086 because it is about the same size, not the other way around.

But I think you may be right about communication between nano devices being a problem.

Linguofreak

Citizen Joe said...

I'm more in favor of the former type of nanomachine where it is designed to do one specific thing.

example 1: Nanites with micro-balloons of expanding foam seek out leaks in the hull of a spaceship. At a certain vacuum level, their shells rupture and the expanding foam fills the void. Additional chemical reactions within the foam may make the hole obvious for damage assessment.

example 2: bio-nanites could be injected which then follow the neurological pathways and then grab on. They can then pick up the neuron messages and pass them back through the nanite chain to a multiplexer (probably outside the body) which analyzes the signals and remotely controls some device. First application could be for paraplegics. Advances in this form of therapy could lead to a 'phantom limb' situation where the patient is able to control a never-existed limb, like a cellphone.

In both cases, the nanites work very simply to move to a specific location and wait there. Technology then takes advantage of their physical make up to do other things.

Jean Remy said...

I must be getting old, I never even thought of a laptop in my earlier argument. I can also blame the time (3 am). Either way, wow, time to take a break for me huh?

Anyways, I am now going to attack the problem for another angle:

The problem with the whole attaching to neurons thing is that a single neural "chain" is generally not sufficient to convey all the message for a given message. Our nervous system is far more analog than digital, and depends on a neural "tree" rather than a chain.

Take pain. Say you poke yourself with a needle (to avoid clouding the debate with a large area to start with. Even that needle point will trigger several receptors, but even let's focus on Neuron 0, the sole receptor triggered.) Let's call the message Ow. Message Ow arrives at Neuron 0's soma, which divides into dendrites. Each dendrite receives Ow which then produces a quantity of neurotransmitters that will convey Ow across the synaptic gaps. Neurons 1, 2 and 3 receive the message Ow. The amount of the neurotransmitter is sufficient to trigger Neuron 1's acceptance of Ow and passes it around. However in Neuron 2 and 3 the amount of neurotransmitters is below a certain threshold. Instead of transmitting Ow, Neuron 2 decides to fold and does not transmit the message. In Neuron 3 the level of neurotransmitter is below another threshold. Instead of Ow, Neuron 3 creates a new message, Not-Ow, an inhibitor signal that now is going to travel in parallel with Ow. If Ow and Not-Ow reach a Neuron 4, the difference of intensity between the two messages will more or less dictate which gets forwarded. Once you get to the pain receptors of the brain, a tally is made of how many Ow and Not-Ow have arrived, and the difference is translated into the amount of pain that was initially received.

For or nanobots to get a real idea of how our nervous system work and we still don't understand some critical aspects of it. We would need as many nanobots as there are neurons to get an accurate picture since a nanobot attached only to Neuron 1 gives a woefully incomplete picture. At worst, it depends on how many neural connections there are (ie: dendritic gaps) since that's where the real work is done.

I hope those things are cheap: you're going to need a lot of them.

Oh one last point: although in diagrams there are neat "zones" in the brain that divide up the work, in reality those zones are more like saying "ok about 90% of this activity type happens here" but there is little to no physical differences between the neurons. Patients that had to undergo partial lobotomies have been studied, and the zones of responsibility are quite free to migrate to non-damaged parts of the brain and resume function.

I think that very soon we might hit a plateau (and indeed may have already) in terms of our understanding how the brain and nervous system works, so the technological bottleneck might not even be processing power miniaturization, but simply that we just don't know where to plug all this neat technology into the brain.

TreverSlyFox said...

Jean Remy said: "...but simply that we just don't know where to plug all this neat technology into the brain."


I have to agree with this at least in any "near future" and anything less than 100 years I consider the "near future".

Right now there is just too much we don't know and not just about the human brain. Though our technology accelerates as each new tech is discovered it's still a slow process because not ALL technology advances at the same speed.

Just the same as while our computing power doubles about every 18 months our "Storage Media" lags well behind that. We may have that Big Bang processing chip in 25 years. But it's questionable if we'll have the storage media with enough space and speed of access to make full use of the processing chip we have.

Anonymous said...

I think cooling in the body is probably easier than in the air. There is constantly fluid flow in the human body, because fresh nutrients must be delivered and waste products removed. Fortunately for the engineers involved in cell-sized technology, that fluid is largely water, so it has a great heat capacity. I think it will only be very serious machines that could produce more heat than the human body could get rid of.

On the subject of communications. The easy and low-tech way (copied from bacteria, like I imagine all near-term nano-tech will be) would be to release a signal molecule into the local neighborhood. That signal molecule gives information on the nano-bot's status to partner nano-bots and allows them to alter function accordingly. There are, of course, serious limitations on the weight of data that can be conveyed this way. It would only work between specifically designed devices.

Michael

Anonymous said...

Citizen Joe said:

"example 2: bio-nanites could be injected which then follow the neurological pathways and then grab on. They can then pick up the neuron messages and pass them back through the nanite chain to a multiplexer (probably outside the body) which analyzes the signals and remotely controls some device. First application could be for paraplegics. Advances in this form of therapy could lead to a 'phantom limb' situation where the patient is able to control a never-existed limb, like a cellphone."

This could be a use of nanites, but the same thing is being developed right now using conventional electrodes and wires. I suppose nanites would allow transition from an "in-patient" to "out-patient" procedure.

Michael

Rick said...

Wow, the comments are proliferating like, well, nanites!

Sabersonic - You raise an interesting oblique point. Clarke made his famous statement, 'indistinguishable from magic,' before the Tolkien era, and since then magic has pretty much followed Moore's Law. :-)

Ian - Giant robots have obvious entertainment value, and so in a shuddery way do creepy crawly bug sized robots. Mere mini robots probably have more practical application than either, but have trouble passing the Hollywood mental threshold that even book writers struggle to avoid.

Jean - based solely on statistical data of relative progress over the last 50 years or so, and even then seemingly only in the matter of information technology ... Very much the latter! If you looked at aviation you'd judge that the age of dramatic progress has ended.

Modern airliners have a lot of subtle refinements over the 707 (especially in information tech) but their basic performance is the same. Compare that to IT in the same era, or aviation in the preceding 50 years, when performance roughly doubled every decade. A lot of technofuturism makes very linear assumptions.


And on the main subthread, I agree that we understand very little about how the human mind-brain system works. In fact I think the biggest single thing we've learned from AI research is how different the human mind is from the stuff we do with computers.

Remember when they thought that only an 'intelligent' computer could play decent chess? Chess computers don't think like grandmasters. They just number crunch enough to do what no human grandmaster can, take the equivalent of a few billion years to play out every possible variation for multiple moves ahead. I think of this as a mental lever, an implementation of human intelligence rather than a display of machine intelligence.

The same applies to the Internet. It has had huge impact, but in a way it just does very simpleminded things. Basically it just combines word processing and document storage with an electronic teletype system. It is what we do with it that is remarkable and unexpected.

We could suddenly discover how the human mind works next year, for all I know. But there is no particular evidence that the Industrial Revolution has seen exceptional progress in human self-understanding. Freud came up with one big insight, the subconscious, but on the whole psychology as a discipline is a lot like AI research - it has told us lots of cool stuff, but hasn't provided what was once expected, a master explanation of how we tick.

So there's no particular reason to think that we'll have figured out human thought by 2100, or generally in the midfuture.

Jean Remy said...

I'm not entirely sure cooling the body is easy.

Raise your body temperature by one degree and it's called a fever, and you can definitely feel that and it's not comfortable. Our bodies do not have a lot of redundancy in heat transfer. We're pretty much optimized to dump the heat we generate now. The nanobots will not be able to dump the heat anywhere but in us. Multiply by the probably awesome number of nanobots, and enjoy a nice slow-roasted-from-the-inside human. The Other-Other White Meat.

Anonymous said...

Great topic!
"Nanotech of the gray goo type seems to have had a brief period of vogue in SF in the years around the turn of the third millenium CE. After a short time, however, the nanotech subgenre vanished, leaving no identifiable traces. It is speculated that all such works, along with their authors, were turned into gray goo."
Obviously, like a well designed nanotech machine, it was self-limiting...

Sabersonic-"The only true logical usage of nanotechnology such as its mainstay of nanomachines are through medical and computing use. Everything else is simply too grand a scale for the passe tech to perform and would require a sense of "magic" to perform. And no, I don't mean the quote about "Any sufficiently advanced technology can be interpreted as magic", I mean Dungeons and Dragon, Lord of the Rings kind of magic."--I've heard arguments that nanomachines may be too fragile to survive the rough handling that exposure to the world-outside-of-the-lab or out-side-the-body would present; so, no nanite-filled artiliry shells, no ultra-fast gray goo producing nanites, and no sprayed-on nanite clouds...medical and industrial nanobots are probably ok, but, while cool, they may not be much more effective than conventional tech.
Rick-
We use medium sized robots now, so that trend will probably continue...I feel that Jean is on the right track about a size threshold for practical human-used devices.

Micheal--"On the subject of communications. The easy and low-tech way (copied from bacteria, like I imagine all near-term nano-tech will be) would be to release a signal molecule into the local neighborhood. That signal molecule gives information on the nano-bot's status to partner nano-bots and allows them to alter function accordingly. There are, of course, serious limitations on the weight of data that can be conveyed this way. It would only work between specifically designed devices." Releasing a signal molecule would also limit the number of times the nanobot could signal the next one, or it would have to incorporate some method of replinshing its store of signal molecules, making it more like a real bacteria than a machine.

My personal opinion is that nanotech is cool, but that it (like any other tech), is best-suited to certain applications and not as an end-all to all facets of everyday life. We are just starting to figure out what this new tech is best suited for, but we're not there yet...kind of like in the middle of the 1800s people envisioned steampowered devices would run every technological aspect of everyday life...as we all know, that Victorian vision of the future didn't turn out the way they thought it would...and, I suspect, neither will nanotech for us.

Ferrell

Jim Baerg said...

"Our bodies do not have a lot of redundancy in heat transfer. We're pretty much optimized to dump the heat we generate now. The nanobots will not be able to dump the heat anywhere but in us."

Maybe nanobots will only be popular with people living in cold climates & we will set the thermostat in the 10-15° rather than 20-25° range;^) Also eat more to provide the nanobots with fuel.

Jim Baerg
51°N 114°W

However, summer might be a problem. The present outside temperature is 24°

Citizen Joe said...

RE: Nanites in biomedical and neurological fields.

Yes, I did base the nanite interface off the current probe techniques being used for limb replacements. While I won't say that it is easy, the principle is fairly straight forward. Create the network of nanites. Get the patient to 'will' his leg to move. Record the neurological map. Once that neurological map becomes stable, i.e. he wills it the same way consistently, feed that data to a robotic arm causing it to move the way he wanted it to. This means years and decades of effort and relearning how to do even the most basic skills, but it can give you back functionality.

The drawback is the very real danger than the nanite network will short circuit the brain and cause seizures. Also, the nanites may very well be quite sensitive to magnetic fields, which could result in the brain being shredded if you pass too close to a strong field.

Jean Remy said...

Citizen Joe: "[...]which could result in the brain being shredded if you pass too close to a strong field."

Is that what they mean by "Rapture of the Nerds" ?

Kedamono said...

@Jean Remy
"Is that what they mean by "Rapture of the Nerds" ?"

I think you mean the "Rupture of the Nerds". :-)

I actually met Drexler in the 80's when he was first touting nanomachines. However, at the time he pointed out that a 1 inch cube nanocomputer would take in near freezing water in one end and vent live steam from the other if it operated at then computer speeds.

We tend to forget that nanomachines are made up of fairly unnatural shapes that may or may not be stable. They also produce heat that could make them more unstable.

In the Dark Horse comic "Dirty Pair: Sim Hell" the antagonist had trillions of nanomachines in this body, rebuilding and repairing it at superhuman speed. Unfortunately, that generated a corresponding heat overload. Our erstwhile protagonists, Kei and Yuri, realize this and beat the Hell out of the bad guy, making him cook in his own juices as a nanites baked him from the inside while at the same time repairing the damage done to him both by the Dirty Pair and themselves. Nasty.

Nanomachines are fairy dust as depicted by most writers. They have no clue about realistic problems with such a technology, such as communication, fuel, and information processing.

Marktheother said...

Nanotech seems best suited to materials science, the assembly of small complex structures, and only the most basic free floating devices (like say little capsules that chemically detect high blood sugar and release their load of insulin)

Some of the stuff that stories use nanites for might be possible for more robust micro-scale robots, not the "magic" stuff but like sensor dust, or breaking down large soft structures (like organic waste).

for communication on the nano scale reception's pretty good with the nanotube radio

http://www.physics.berkeley.edu/research/zettl/projects/nanoradio/radio.html

but I think there are some size limits on transmission, but I'm not sure.

-Mark

Rick said...

I wonder if real nanotechnology will even go mainly in the direction of 'nanites,' or instead (as is already happening) toward things like 'smart materials' with whole layers of embedded nanomaterials.

The problem with nanites isn't that they turn everything into gray goo, it is that if you drop one you have the devil's own time finding it. Putting them in the human body, for example, and getting them to go where you want them to go is a big and complex navigation problem! An array of them embedded in something like foil is a lot easier to work with.

Citizen Joe said...

You can also do bio-nanites. You piggyback your manufactured nanites on a bacteria and then let the bacteria go to where it normally goes. Once there, you activate the nanites with some sort of EM pulse.

There's all sorts of ways to get them to their locations, but again, don't go for the intelligent/complicated nanites, just something simple, almost just a tracer type nanite.

Jean Remy said...

Well if you're going to use genetically-engineered bacteria, why not use engineered retro-viruses? I guess retro-viruses could be considered biological nanites (but not necessarily. Your millage may vary). In this case however once they are "programmed" you can't really reprogram them with an external signal, you'd have to inject a new batch of viruses to replace the earlier ones. The problem with viruses is we don't know how to fight them (yet) so we'd better not make a mistake in the initial RNA code.

Either way, we have not yet managed to build a fool-proof computer, and nanobots or retro-viruses would be hard to destroy if we do make a mistake since I don't foresee a way to damage them without damaging the host.

Really, I think that's my main objection to any semi-independent technology released inside a human body. Releasing them without a way to stop them without damaging the host would be utterly irresponsible.

I certainly don't think the next 25 years will invalidate Murphy, and I'm not about to put money down against him.

Anonymous said...

Joe and Remy touched on my thoughts about nanotech: Whatever we build will look an awful lot like bacteria and viruses, and will probably even be built out of bacteria and viruses. We already know that they work (Too well, in some cases. Have you ever tried to scrub a biofilm off your bathroom wall?), now we have to figure out how to make them work for us.

Actually, our record is pretty good on that last part. We discovered yogurt, bread, cheese, and beer with absolutely no knowledge of genetics. It should be interesting to see what we do with practical genetic engineering. Anyone for yobeer?

Ian_M

Rick said...

Betting on Murphy is historically the safest of bets.

M. D. Van Norman said...

How about nano-devices as the “cells” of a living machine, such as an advanced automated spacecraft? Robots that can effectively modify themselves and/or “heal” on the fly?

That said, most nano-tech science fiction I’ve read seems to completely ignore the vast amount of waste heat these critters would put out while rapidly assembling consumer goods and other widgets.

Citizen Joe said...

Nanite 'cells' may be how buckytubes are grown.

Anonymous said...

Jean Remy said:

"I'm not entirely sure cooling the body is easy.

Raise your body temperature by one degree and it's called a fever, and you can definitely feel that and it's not comfortable. Our bodies do not have a lot of redundancy in heat transfer. We're pretty much optimized to dump the heat we generate now. The nanobots will not be able to dump the heat anywhere but in us. Multiply by the probably awesome number of nanobots, and enjoy a nice slow-roasted-from-the-inside human. The Other-Other White Meat."

I think you're wrong. The human body has limited methods of temperature mitigation (bloodflow and sweat) but they are massively redundant, because you have hundreds of sweat glands and blood flow is necessarily reliable. Skin can also change its insulative properties. Humans can be sedentary or exercise, from actic wastes to tropical rain forests (more limited if you don't take into account clothing) and manage to keep the core temperature within tolerances.

Fevers aren't accidental increases, they are purposeful increases. The body is actively warming itself. They are not due to a lack of cooling.

Citizen Joe said...

Ironically, the defense mechanism of fevers use heat to kill viruses and microbes. That same mechanism would likely fry nanites.

Along those similarities, viruses would be an excellent model for nanites. They have a very simple chemical signature seeking program. If they don't find that chemical signature, they don't latch on and just keep floating around until they do find it. They don't actually 'do' anything, but rather are carriers for a genetic program that they pass on to more complicated cells.

Jean Remy said...

You are absolutely correct. A fever is a change in the body's thermoregulatory set-point. I should've said hyperthermia which involves heating the body past the set-point.

I also never meant to say that a fever was involuntary, either. But internal temperatures during a fever are far from reaching boiling point, which is the recommended temperature to sterilize any unwanted organisms. A rise of a couple of degrees might not be as effective as you think in actually killing the organisms. Whether the body uses the heat to kill foreign organisms or as a means to accelerate the immune system's response is still rather misunderstood and controversial.

Whether we're taking about hyperthermia or fever, and whether it is involuntary or not, my argument still stands: it does not take a great internal change in temperature to start causing trouble. I agree there are many thermoregulatory tricks to combat external changes in temperature allowing for humans tom live in wildly different climates. However even in wildly differing climates, internal mean temperature is between 37 and 38 C. The means to keep that temperature constant (from sweating to shivering) are not necessarily comfortable.

Although nanobot waste heat generation could be handled if the number is limited, that goes against the point of nanomachines. Enough of them will generate amounts of waste heat that the body is not designed to handle, and would cause hyperthermia.

Under monitored and tightly controlled circumstances I can see nanobots used. However I doubt they'll be used so widely and offhandedly as they've been portrayed, certainly not to the point where humans are more cyborg than flesh, walking around with billions of them futzing along our nervous systems.

Rick said...

I don't know much about the mechanism of fevers, but a body temperature a few degrees above normal will kill you if it persists - I'd bet on heat generation to kill a person before it kills nanites.

But I'm rather doubtful of a big role for independent nanites, as distinct from nanostructures attached to something larger.

Anonymous said...

Let's take a person with 50 kg of water (a ~56 kg person, on the smallish side for an American).

50 kg * 1000 g/kg * 1 cal/degree/g * 4.184 joules/cal = 209 kJ of waste heat. So you need 209 kJ of waste heat beyond what your body can easily reject to raise it 1 degree C. Call it half that for 1 degree F. I don't know heat rejection rates of the body, so this value isn't that useful, but it's something to consider.

I do know that people have jobs where they perform manual labor 10 hours a day in the summer in humid 80-90 degree F heat. Historically, people have worked in pretty hot mines for long shifts, also on farms, in factories, etc. The human body has heat rejection hardware designed to handle high heat loads for long periods of time. I can't imagine any array of nanomachines that would have a function that required even 10% of the energy required to drive railroad spikes all day.

This comparison is not completely fair, because one doesn't want nanomachines to cause the user to sweat profusely while doing paperwork, but the comparison holds.

Nevertheless, consider that a human being can easily survive nude between 70 and 100 F (while living a physically demanding life), and you realize that a little extra heat is trivial. Think up a nanomachine that might use power on the order of human muscles or organs and I'll consider that it might cause heat load problems.

Michael

Rick said...

Michael - A few additional data points: In rowing, which surely qualifies as heavy labor, a person can generate about 100 watts of actual propulsive power for a short time, maybe 40 watts sustained. Work done by the body is more than that, and waste heat several times as much. So I'd guess that a person working hard all day is generating maybe 250-350 watts of heat. Sweat is the primary coolant; water consumption aboard galleys was enormous and a primary constraint on their operating radius.

But dumping heat from internal nanites could also depend on where they are - I imagine the body's internal heat management capacity is focused on tissues that do a lot of work and therefore generate a lot of heat. A part of the body might get dangerously overheated even if the body as a whole would be able to shed that amount of heat.

Anonymous said...

Rick, I agree with you, and thankfully that allows for quite a bit of real estate. Skeletal muscles are large and plentiful, and organs are also highly vascular.

This, of course, leaves out the brain. If you don't want to risk upsetting brain thermal balance (understandable), you can probably implant a lot of brain pertinent nanites into the skull, or simply have heat conducting wires lead from the device, through the skull, to the plentiful blood flow around the skull.

I want to restate the point that your working fluid is effectively water, and that is GREAT for cooling.

Michael

Rick said...

Water is a great coolant, and sweating absorbs something like 600 Kcal/ltr of waste heat.

All that said, I'm not quite sure what swarms of nanites inside the human body would actually DO.