Sci-fi Military Tactics and Strategy

You wouldnt have a reference on the rate of copy degeneration for a given technology, would you?

Probably not mostly because no one made them yet to measure it, unless you hardwire some really strong checksums or even more error correction codes every time a nanomachine replicates itself you could always model the degeneration after bacteria.

Of course unlike bacteria nanomachines may have some utility issues after replication, specially if said degeneration messes up the replication model.

I think a machine or something supplying nano-machines as they wear out is more feasible. A constant master copy approach where it isn't a copy making a copy but a single machine making copies of a single blue print.

There will still likely be errors but it is unlikely wind up with a copy of a copy problem.

In terms of raw mass production you are more likely to be able to sustain your supply of nano-machines without cascading transcription errors possibly sidelining the whole production run.

In fact I think the whole "grey goo" scenario is almost impossible because it require an amazingly resilient self regulating mechanism that is itself almost immune to degradation. That and it sound suspiciously like perpetual motion machine kind of capability. However that doesn't mean you couldn't say use it as a weapon.

Instead of thinking of it as a machine weapon think of it a biological weapon proxy only it infects matter instead of just living organisms. In terms of biological warfare for a weapon to be useful as a precursor to an assault it needs to have a finite ability to spread and will effectively burn itself out. This is usually achieved with rapid onset of symptoms and a high lethality rate.

Only instead of that these machines have an estimated degradation rate that they reach a certain point and the whole production line simply breaks down from the compounded errors. So instead of a run away chain reaction you can have an estimated end of activity from the machines as they finally reach a point where they can no longer self replicate reliably and the nano-machine population burns out.

Basically even the more reliable initial machines are likely to have a limited life span and if you have a chain of break downs by the time the production hits a road block there is a pretty good chance the originals are no longer able to sustain the chain of manufacture from simple wear and tear and even be contributing to the collapse as they wear out and begin to create inferior copies. Which only accelerates the break down rate as it would increase the error rate from the top end as well.

edited 8th Dec '16 5:17:44 PM by TuefelHundenIV

In organic evolution, recombination processes (like sexual reproduction) are used to preserve genetic information despite a steady rate of mutation. In addition, the simpler the pattern to be preserved, the less susceptible the process of replication is to transcription error. Nanomachines could be made individually very simple, since all they are supposed to do is facilitate some chemical process in the air, sea or soil. Their simplicity extends their resilience to transcription errors, and certain evolutionary algorithms, which simulate the effects of natural selection for programming code, could also help extend the useful lifetime of the swarm. In fact, if the swarm adapts and evolves in ways similar to organic life, I don't see why they cant replicate nearly forever.

This.

Repeat after me, organic life is not special.

Immy: Biological life is actually pretty special given it is amazingly difficult to replicate. Even worse as far as we can tell it requires rather specific conditions in which to even appear. So yeah actually it is pretty special until we find out otherwise. We have done more with imitating it and simulating it then anything else.

De Marquis: Except even biological organisms can quite easily develop a flaw or error that without outside intervention would doom them. It has happened from simple to complex organisms. Bacteria colonies arguably some of the most simple organisms on earth can and do easily develop mutations that cause a colony to die off even inside what should be an ideal environment. Being an organism that can mutate or evolve does not protect them from transcription errors by any stretch of the imagination.

It only takes one error to cause a cascade of problems down the line including failure to reproduce. DNA and RNA are very susceptible to outside influences causing constant errors and something as simple as even a minor change in environment can lead to destructive changes across the whole population. The faster the reproduction cycle the faster it happens. Making a nano-machine like a bacteria doesn't save it from rapidly piling on transcription errors. Even pointing to evolution you can just as easily point out the fact even simple organisms can go extinct so no they can't replicate forever and are far from immune to transcription errors especially if you model them on organic life. Evolution can just easily pass on negative traits as positive ones and isn't any more protection then any of the other factors. The problems of trying to copy yourself catch up pretty quickly when your generations happen very quickly. The faster they are the faster it catches up.

There is also another problem you didn't touch with evolving machines. That is they can evolve to do something you don't want them to do or even stop doing what you want them to. That also introduces more chances for negative trait to perpetuate it self. Not all evolution results in successful long term organisms and again evolution can pass on a negative trait just as easily as it can pass on a positive one. The only way to stop that from happening is to limit their ability to adapt and change outside of certain parameters.

A nano-machine is quite unlikely to have a very long life span to begin with and the more work they do the more likely they are to break down just like any other tool or machine. When you talk about using them to cause large scale terraforming you are not doing a small amount of work.

Altering an environment is not a simple task by any measure and is not a matter of simply altering a few chemicals . The closest you get simple in that regard is if the nano-machines are simply depositing a substance into environment. Even then that has to be carefully controlled and is only part of the bigger picture. Terraforming as we are discussing it is not a simple matter of altering one aspect of an environment but altering the whole environment. That is the air, water, soil, microbial life, and even complex organisms.

No nano-machines are ultimately limited by the mere fact they cannot escape transcription errors and attrition just like biological organisms can't. The faster the nano-machines reproduce the faster those errors will pile up and reach a point where it notably affects their ability to reproduce.

I have never heard of a species that went extinct due to DNA transcription errors or random mutation. Got a link, please?

The problem is that there is no magical mumbo jumbo, no impossible to replicate thing about organic life.

Difficulty doesn't mater so much here, since we are talking science fiction, and from a physics, chemistry, and engineering standpoint there is nothing about biological that is impossible to replicate.

So "incredibly difficult" isn't an argument, besides were approaching the point where we can replicate it rather fast even now.

Immy: Your conflating imitation with being the same. It simply isn't. Which leaves life as still rather special until we can do or make structures in the same way. To date we really can't. We do a lot with existing samples encouraging what they do naturally to create something else like growing organs. We do even better when we alter existing forms modifying them to new tasks. Bio-mimicry is just that mimicry not the real deal, not the same. We are still puzzling out insect and bird flight something humans have been studying for a few centuries now and even now have some limited success to imitating it but the real deal is still superior in many ways. It is quite a bit harder to even mimic it then you make it out to be. Which is why bio-mimicry is intensely studied and also partly why groups like DARPA said screw it and created cybernetic bugs. Altering an existing form seems to be quite a bit easier then making one from scratch.

De Marquis: Becoming sterile which is a common result of various hybrids. That is a non-self sustaining species and has to be bred from the animals that made the hybrid. Becoming vulnerable to something the in the environment that they used to live in which is the most often cited example of harmful mutation. Constant inbreeding of Cheetahs is killing them off from a whole host of health problems. All of those could be considered a transcription error in DNA. Inbreeding is one good way to fast track that in complex animals like mammals.

And here is your scholarly link Mutation Induced Extinction in Finite Populations: Lethal Mutagenesis and Lethal Isolation They do point out that in the real world effectively all populations are finite. Later on in the article they point out the rate of mutation will eventually kill off even theoretically infinite populations. Even better a lot of the cited sources are for study and theory on bacteria and viruses.

There is a lot more to it where they explain how they are modeling everything and various bits of information come from.

Basically it boils down to this. Everything will eventually mutate to the point the organism cannot survive sans intervention or lucky correction of detrimental mutations. The faster a species reproduces the faster this happens.

Now with that article in mind you might get more mileage out of the nano machines by making them less like organisms.

This is also part of the allure of trans-humanism. With technology we could go in a completely different direction or even multiple different directions in theory.

Unless a substantial portion of the population picks up the same mutation (perhaps due to The Virus), lethal mutations are more likely to prevent themselves from spreading further because, well, they're lethal (this of course depends on whether the mutation kills the organism before it reproduces or after. Evolution doesn't select for the better being, it simply selects against the one that can't reproduce)

AFP: Not actually true. You, me, and everyone else in this thread are a walking batch of mutations. They will give us things like cancer, make us vulnerable to various diseases, and even lead to our bodies breaking down in rather unpleasant ways. All a mutation has to do lead to termination of a population is cause the populations attrition rate to outstrip its reproduction rate. Again making organisms more vulnerable to disease, environment, and even reduced reproductive success are all recognized factors leading to a population dying off. The number of sources that can inflict mutation on even a fully developed being are numerous. Every last one of us has likely developed more then one mutation and are walking around with it in our genes that will quite likely lead to our deaths.

Again a bigger factor is rate of reproduction and as the paper notes mutation is pretty much inevitable and has a startlingly high frequency. A frequency higher then even I initially thought. The faster an organism reproduces the faster the rate of mutations pile up to the point something bad starts happening to the population. In humans we have also gotten to the point where deliberate intervention is blunting some of those detrimental effects extending our life spans and ensuring we can successfully reproduce. It also helps ensure or rate of reproduction remains higher overall then our rate or attrition. Sans deliberate intervention human mortality rates would be quite a bit higher. We have effectively more then doubled our average life expectancy from the earliest estimates thanks to advances in medicines and tools allowing us to intervene in the first place.

By the earliest standards they have an estimate for you, me, De Marquis, and likely several others would be "long in the tooth" instead of middle aged.

edited 9th Dec '16 7:39:50 PM by TuefelHundenIV

@Tuefel: There are several problems with that article, but the main one is that they do not come to the conclusion you claimed for them. From the article: "Thus, we conclude that stably folded proteins are only possible in ecological settings that support sufficiently large populations." In other words, if the population is large enough that the reproduction rate exceeds the mutation rate, the population can continue to exist essentially forever. Which should be self-evident: after billions of years, life on Earth still exists. By "all finite populations eventually go extinct" they mean that statistically speaking eventually a generation will come to pass in which a majority of members will simultaneously experience a fatal mutation, leaving too few survivors to re-populate the environment. This may or may not be true (genetic evidence of multiple human population "bottlenecks" argues against it), but in any case they give no estimate of the number of years this might require. I think it's self evident that for any population of a reasonable size, the length of time required for this to happen purely by chance would be measured in billions of years- which means that while my nano-bot swarm may not survive until the heat-death of the universe, they should function long enough to transform the planet's environment in a significant way, which is all we need them to do.

"All a mutation has to do lead to termination of a population is cause the populations attrition rate to outstrip its reproduction rate." This assumes that nearly every member of a species receives a fatal mutation within a generation (otherwise the survivors eventually repopulate the environment, exactly as AFP says). The rate of lethal mutations may be quite high (for a given value of "high"), but they wont spread because they undermine the reproduction fitness of the individuals who possess them. So they must occur across a majority of the population all at the same time in order to bring the reproduction rate below the mutation rate, and this must occur entirely by chance.

However, the main take-away that I get from all that is if you want to seed a planet with a self-reproducing swarm of nano-bots, make sure that their fabrication rate exceeds the rate of transcription error in the code. Another implication is that the planet's inhabitants might defend themselves if they could find a way to increase the probability of transcription errors- perhaps by means of an electro-magnetic field of some kind.

"By the earliest standards they have an estimate for you, me, De Marquis, and likely several others would be "long in the tooth" instead of middle aged."

Heh. What makes you think I'm middle-aged

edited 11th Dec '16 11:08:02 AM by DeMarquis

De Marquis: There are not any problems with it and they explain exactly what I am getting it. It details how they built their calcuation model and cite their data sources. They also state that with both finite and infinite populations as I already noted, that ultimately populations are hurt sufficiently by mutation that they can die off based on their modeling. Try again.

This showed up in my youtube feeds today. A movie about a rogue experimental AI and all of its kill the humans goodness on a small scale. Also has cyborgs and hacking and AI and Cyborgs ineracting.

edited 11th Dec '16 3:56:59 PM by TuefelHundenIV

I remain unconvinced that transcription errors prevent self replicating devices from achieving significant numbers.

The best fictional depiction of AI that I have seen remains the webcomic "Freefall".

edited 11th Dec '16 7:12:43 PM by DeMarquis

It's not really the transcription errors so much as there's no good way to fix these errors. You can have built in backups but that adds bulk and on something the size of a nanomachine bulk defeats the purpose. Even then, it's more of a stalling tactic as there's no reason the backups can't be corrupted as well. Wear and tear then takes over with existing examples and the entire thing falls apart.

Organic life solves this by recombining viable lines into new lines. Sex, basically. In this way we can control copy degradation by selecting out the fatal errors and using eachother for parity data. Mutation is inevitable but fatal mutation is pruned.

Thing is, this system inevitably leads to controlled mutation if only for how it selects for the best at self-replication. This makes it kind of useless as a tool since you never know what it's going to do next.

They do mention that if you can effectively preserve an original master copy the population could in theory survive indefinitely. The problem is virus and bacteria do not do that and neither does RNA or DNA, instead they have two copies with at least some inherent mutation that occurs during replication.

Bel: Sex does not control degradation at all and helps spread it. The Hapsburg line for example died out because it couldn't stop degradation and the last of them was effectively unable to procreate. They were massively inbred to a horrifying degree. Sex only works with sufficient differences even then you still get fatal mutations.

For a change of tack otherwise we will be at this forever.

Soft Sci-fi stuff this time. If you had a walker of any number of legs what would you try and do with it? What tech might you pair it with?

For me something that can climb vertical or negative slope surfaces using the gecko like material they have been working on and carry several pounds of equipment at the same time. That way you can use it to climb just about any surface and haul around sensing and tracking kit or even small weapons packages to attack or spot from unusual angles.

edited 12th Dec '16 4:15:00 PM by TuefelHundenIV

I would go with 6, that's stable, has built in redundancy so you can loose a couple, but isn't getting to the point of absurd complexity yet.

As for tech pairing, either automate it as a drone, or make sure that at-least the turret is.

edited 12th Dec '16 3:40:46 AM by Imca

All right people, tell me about your Space!! Special Forces and what are there roles in this era of Space Warfare.

I figure you'd have a Starship Troopers esque drop trooper element. I also figure that trying to insert them into anything like contested airspace would be extremely risky, so they mostly use it for relatively discrete (like a Drop Ship would be too easily noticed but some cap troopers could be made to look like debris or something) or very time-sensitive operations.

You could use the same Drop Pods to do supply drops, in which case the contested airspace is less of a concern because if a pod full of ammo gets blown up, it's the same difference to the guys on the ground as if you didn't try to supply them at all.

For when you do have to risk the It's Raining Men approach, you'd also have ECM, decoy pods (to include supply pods dropped with the troopers), and weapons pods designed to target and destroy enemy air defenses (as well as to intercept incoming fire). The whole operation would of course be quite expensive and dramatic, meaning it would have to happen at the climax of the first book at least.

edited 14th Dec '16 6:37:05 AM by AFP

I'm pretty sure if you simply keep the fabrication rate high, and insert several thousand at one go, the swarm can keep itself going for some considerable time.

I agree that for a very large mech-like vehicle, 6 is the minimum. Its a similar consideration as for the number of wheels on an armored car- the more legs, the less the ground pressure per leg, and less catastrophic the loss of one leg is.

I'm going to assume that the primary mission of Deep-Space special forces is capture of enemy facilities and equipment. They will need the capability for long-endurance suits, a way of cutting into a ship or habitat safely, and specialized hacking equipment.

edited 14th Dec '16 3:18:28 PM by DeMarquis

De Marquis: Or even periodic injections of a fresh population if you want to keep it going long after the initial ones are gone.

The role of special forces isn't likely to change hugely so much as the environments they have to operate in. There is a cheesy and short lived pulpy book series that explores that with Navy SEALS taking on the role of space borne troops as well. They have a new combat space suit, new weapons for vacuum, and a bunch of other stuff including procedures for breaching into a sealed habitat.

You also need a means to move them around either in vacuum of space, on surfaces likes mars or the moon, or even to transport them through atmosphere and moving around on Earth like worlds.

What about boarding Spec Ops for infiltrating ships and space stations? Or should that be a skill taught to every Space Marine, since it's more than likely a Marine would take part in Zero-G training

I just had an amusing idea of shipboarding being a skill mostly limited to special ops types and the Coast Guard In Space

What would you call a Coast Guard in space, anyways? Orbital Patrol? The Revenue Cutter Service? Space Police?

Tactical: Given Marines have assisted in customs and duty inspection as well as Navy the training might be more wide spread.

^^ Coast Guard: Space Division.