![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
beccaelizabethI know, math isn't very interesting. I could just lock all these, but then I can only read them signed in, which matters if I'm pondering at college.
But I've been thinking about all the rules tweaks I chose for my Giant Capertiller Gestalt Matrix Computer.
To start with, in the usual rules for a gestalt, if you plug one brain it, it becomes a complexity 9 computer capable of running 2 complexity 9 programs, and a mind emulation is complexity 9. So basically you can run (a) yourself and (b) any actual computing you want to do, both in the same body. That makes logical sense.
But you plug 8 more bodies in and you can still only run one mind in them, which doesn't make sense.
You get a lot more work out of them if you pretend they are separate computers that talk to each other.
But to do that you'd need 500lbs of computer each instead of just 50lbs of connector.
... I haven't decided how much the Giant Capertiller weighs. If it weighs 550lbs per mind, it could do that. But that's like four times as much as the humans it is plugged in to. Heavy.
Plus you wouldn't get one shared VR that way, you'd get many that talk to each other. I'd have to decide on data rates and suchlike.
It's unwieldy, complex, and stupid. I will not do that.
So you can have 9 bodies plugged in and still only run one mind. One mind of average intelligence. It's daft.
So I like the vatbrain biocomputer rules, because then you can run extra minds or very smart minds from the first time you plug in. 20 programs worth of them, in fact.
But when I plug in a regular size capertiller to be a backup brain, in some ways, it makes a lot of sense to call it an AI running on a single brain gestalt computer, because then you can limit its IQ relative to the complexity of the system, and know why it can't run without plugging in. It's a hard drive waiting for a computer. I knew that. Plus a gestalt component mind is unconscious while it works, so built in to that is the idea that the host is unaware while the capertiller is thinking. But with capertillers, I want two and only two minds able to work at any one time. And one of those needs to be unconscious if the capertiller wins. And I don't want to mess around counting how many programs a smart person takes up. And, fairly importantly, it does not weigh 550lbs. Quite a lot less really. But it is also two tech levels more advanced than the gestalt matrix. Advance a tech level, get smarter out of smaller. The tiny ones are the advanced models of the big ones. So I'm trying to figure if the ways it doesn't act like a digital mind outweigh the ways it does. There's no VR in there, it can't create its own world, it can't run programs, and it was never intended to interface with other technology. That's quite a list of not-digital. The only way it is digital is it depends on the complexity of the supporting system to have any IQ at all. For an AI, that's a simple taboo trait, and costs 0. So I just need to make the rule 'IQ cannot exceed that of the host' and stop worrying about making it match the Giant version. It's a different device of a different tech level that did away with many features in the quest to miniaturise. Old mobile phones did not do many of the things a mainframe did, even using the same variety of tech.
So it's not a digital mind, it's just a mind with a taboo trait limiting its intelligence to that of the host. Skills go with it, so it can get very good at a lot of things. A backup brain takes a number of months equal to the hosts IQ to create a full record of the host, one that can just go and run around being the host's brain if it feels the need. A possessing entity does not get that kind of access to a host, so it isn't Possession advantage that does that. And it isn't Extra Life (Copy), at least not for the symbiote. If it doesn't get conscious access to all that data then it's just a special effect on Racial Memory (Passive), and if it does, that's Racial Memory (Active). But the symbiote has to remain in a body for long enough to soak all that stuff up before it becomes part of its own memory. Before that they would have to let go of their host, let them become conscious, and either ask them or use the struggle for possession to wring data out of them. And every such attempt would give the host a chance of regaining control, making it immune to that particular 'tiller. But if the host willingly relinquishes control back to its 'tiller they could, like tok'ra, switch off, unless the host body immune system reacted unpleasantly (their HT exceeded their will and physically rejected the symbiote) or the 'tiller totally bungled the catch (crit/fail its will check for possession). I don't know what a fail would represent in that situation, sticking their neural probe legs in the wrong nerves? The possession equivalent of forgetting your password? Still, it's pretty unlikely once you buy your will up high enough.
But I've been calculating the will for initial possession purposes as if they already had an IQ, the IQ of the host body. They don't get access to that unless they win, in logic world. Hrm, it's like when a virus program loads on a computer it shouldn't be in, it runs as fast as the system allows, right up until the virus checker kicks it out. Similarly, grabbing all those nerves gives the 'tiller access to the hardware right up until the software or the immunities kick it out. There's sufficient logic there.
But if they do that there's more chance to possess a high IQ host, since there's less chance of rolling a fail, than there is of possessing an animal, because the 'tiller in an animal isn't bright enough to figure out what it's doing and might just bungle the plug in. That... is not a bug. They were designed to back up intelligent beings, not to pilot animals. And we wants them to try and borrow people who will know enough to not like it. Drama. That works out.
Okay, so the remaining constraint is, how long does it take for a blending to transfer data?
And I don't know, really. It shouldn't be instant, but it can't take a number of months equal to cumulative IQ of previous hosts, which would be the logical extension of the Backup Brain rule. Perhaps I shall just say that when two 'tillers like each other very much they go back to the sea and Stuff Happens.
A capertiller can split in half, and retain lots of data and some of its will and fatigue. Data is really easy to grow space for.
Data Storage: Additional built-in data storage can be purchased for $1 and 0.001 lb. per additional TB (at TL9). Multiply storage by 1,000 per TL after TL9.
Capertillers do things that are only possible at TL11, like being backup brains, but their data storage could be a lower tech level relic. They're also biogadget, like a vatbrain biocomputer, needing to weigh twice as much because of life support.
So each new bit of data storage weighs 0.002lb. If it's TL9 that's 1TB, TL10 1,000 TB, TL11 1,000,000 TB.
It is not TL11 because that's ridiculous.
One human mind backs up onto 100 TB.
Even at TL9 that's 0.2 lbs of data goop. You could store three minds in a can of soup. If their data storage is more advanced than that... you're just not going to run out of TB. Ever. At all.
In ten thousand years though?
2lb of capertiller is 1,000 TB at TL9. 10 minds. That's only one every thousand years.
But if the data goop is 2lbs of TL10, that's 1,000,000 TB or 10,000 minds, one a year. You wouldn't change hosts once a year. You wouldn't absorb them all in less than ten months unless they were particularly stupid.
So the capertiller has TL10 data storage and needs rather a lot less than 2lbs of goop.
If you say there's one new mind every twenty years for ten thousand years, continuous remembered experience since the cities were lost, that's 500 minds, or 50,000 TB. At TL10 that weighs only 0.1 lbs and we're back down to six to a soup can.
That's just the data storage parts of a capertiller. It still needs a body around them. So the capertiller might even fill the whole soup can.
Which is handy, because, as a parasite, it needs to fit inside a host body, even while remembering lots of things.
I know, I'm comparing weights not density or other useful dimensions. And the GURPS rules just give you size modifiers based on length, they say nothing about weight.
Dwarfism says about weight: Size Modifier -1. Choose your weight from the first line of the Build Table and reduce it by 15%.
The Build Table first line is for height range 4’4”-5’2” and skinny weight 40-80 lbs.
But on the Size Modifier tables 1.5 yards (4.5’) is SM-1, and that's 4' 6". So there's a bit of overlap.
Parasites have to be one SM smaller than the thing they're trying to possess. Which would mean a capertiller can be up to 4' 6" and still curl up inside a regular human.
... eeew?
But it's totally the wrong shape and sort of thing to weigh as much as a dwarf human, even a skinny dwarf, even if it was that long.
The spinal cord weighs approximately 35 g. I googled it. I had not known that before.
The spinal cord is long and skinny and made of nerves, like the remembering bits of a capertiller.
It's also where the capertiller plugs in.
That probably shouldn't work, but then, neither should goopy parasitic backup brains.
The spinal cord is only 45 cm long and then turns into a horse tail of nerves that run down the rest of the bony bit.
I am learning many things today.
Being longer than the thing it plugs in to would do the capertiller no extra good.
Half a yard long is SM-4. Capertillers can be SM-4 and lie along your backbone sticking their legs through the gaps. Or just grab the spinal nerves where they stick out.
I had made a formula where Will is related to Size Modifier, but I don't think I want to make their size as variable as my maths was. So I'll redo that idea.
The sticky outy bits of the back bones are all on the posterior side, so the capertiller stays comfy on the smooth anterior side.
I am learning words and making stuff up.
And yes, most of this is because I look at a goa'uld and just wonder where it all goes. I mean they're huge great squirmy things. There's a scan that shows one wrapped around the bones in the neck, Stargate Wiki page on goa'uld shows it. But there's much more squirmy thing when they're on the outside. And necks are quite busy. Can you really fit an invisible thingy in there? Plus, sticking a bit snake head into your brain just seems more likely to create disruption and permanent damage than possession.
... I'm criticising the plausibility of Stargate. *facepalm*
But I just want to put my parasites somewhere more roomy, with plenty plug in points. They don't need to be in the brain to access the brain. Eyes aren't.
So: 45cm long, weighing at least 0.1 lbs or or 45g, probably more so it can have life support. A Backup Brain is a Tiny Computer, which weighs 0.05 lbs, or 0.1 lbs as biocomputer. That doesn't seem sophisticated enough since even at TL11 it's only Complexity 6, but I don't want it to run a mind emulation, just remember them. It only needs much more basic responses to jump in a host. I set its IQ as 2 the same as a rattlesnake, and you can run an IQ2 volitional AI program (IQ/2)+3= Complexity 4 -1 for biocomputer, so IQ 2 VAI C3 program, in even a TL9 Complexity 3 tiny comp. And it's likely to be TL10, and still smarter. So, more weights and measures:
up to 45cm long, weighing at least 0.2 lbs or 90g. That's only like three times as much as a spinal cord. I probably need to add some weight to be all the little plug in arms. And it needs a power supply, but it gets that by leeching off the host body. It's just a feature of being a biogadget and doesn't have a weight multiplier attached, that's already taken care of by being a biocomputer. So, just the tiny arms, but those are to carry neural interface systems, which have no weight listed. It's not going to weigh more than 0.3 lbs unless I tell it to, basically. 135g, waaaay less than a can of soup. But if it was going to split in half each half would still need to be 0.3 lb so a growing 'tiller needs to be up to 0.6 lb or 300g before it can survive splitting with all its memories intact. If it gets split when its smaller than that it probably loses a lot of data storage. Amnesia time. But if the chunks of an injured 'tiller weigh more than 0.3lb they are in fact two injured capertillers. It is very goopy and can stretch and squish a lot. Its length isn't the important measurement, just its weight... which is unuseful when GURPS size modifiers are all in length and ignore weight.
An SM-4 humanoid, if it weighs 15% less per size modifier, is 18" tall and weighs 20 lbs. Searches say babies don't weigh that much at that size but they can't move around on their own neither. But snakes are more like the right model for a limbless capertiller. A snake that is sometimes 18 inches long can weigh 0.5 lbs or less, though there's loads of kinds of snakes and weights so that's difficult to pull relevant numbers. But the 'tiller weighs like a snake. That's about right then. Though it doesn't have bones or scales, so it should be on the light side.
A capertiller can be up to 18 inches long, and weighs around 0.3 lb, but probably more because it's trying to grow a second one. As soon as it can split up and both halves weigh 0.3 lb then it can survive getting cut in half. It is sneaky like that. It fits along the host body's spine so in a human it has plenty of room, and in most things SM-3 or bigger it doesn't even have to squish itself up much. To fit in a squirrel is trickier.
There's obviously a point past which it's just ridiculous to try and fit. The GURPS rules figure that out by size modifier, but I have instead weight. If one that is ready to split can still be SM-4 then one half that size can get as short as SM-5 and be perfectly healthy. That's like 9 inches long. I have found sorts of squirrels that are that size, 7 to 10 inch SM-5, like the regular red squirrels around here if you don't count their tails, but also there are big ones that are more than 10 inches long, so SM-4. Yes, that's a very big squirrel: an Indian Giant Squirrel, for instance, or Russet Ground Squirrel, also called the Large Toothed Souslik, or the Anatolian Souslik; a zizel is a souslik. But a regular grey squirrel can get up to that big, says wiki. A capertiller being SM-6 could fit into a sort of a squirrel that was SM-4, under the parasitic possession rules. Though it would make it a rather fat squirrel. A single capertiller can possess a squirrel, for sufficiently large values of squirrel.
The only reason they'd deliberately get big enough a human host might not be big enough is if their size lets them grow something useful, like more IQ of their very own, or higher will to possess bigger things. Computers don't grow a complexity until they weigh ten times more. To go from SM-5 9 inch long creature to ten times that makes them 2.5 yards long and SM+1. Humans are 0. They would outgrow a human long before they got any IQ benefit from their increased size. Is sad.
They could try possessing an elephant. There's a lot of room to grow in an elephant.
... Maybe it is best if there are no elephants in this setting.
... or, possibly, you end up with very smart possessed elephants. Huh.
Wait, I might be reading size modifiers wrong. You're supposed to round up, so it's things up to 7 inch or up to 10 inch. Oooh, maths. But then 18 inch is SM-4 but 9 inch is only SM-5. But then all those squirrels who are more than 10 inches long are SM-4. So they can still fit in a squirrel.
I will go up and fix it. Okay, is fixed now, only this paragraph stands as testament to my oops.
If it is not a computer it can get smarter as it gets longer. Then there would be actual point to getting longer, instead of just splitting up. I'd be making a new rule though. I think I'd rather stick to Will boosts rather than actual IQ. Big old capertillers can hold on in more places and that makes them better at possessing stuff. As long as the stuff is big enough. In small dumb things the host can't fight back mentally very well, but the capertiller can't borrow enough brain power to figure out where to plug in, so it's quite likely to fail at possession. In smart things the host can fight back mentally with a very high chance of success, but the capertiller is now borrowing smarts enough to keep up, and has a much higher chance of not just screwing up the attempt.
Oh, I've thought of a reason to get big: if a 0.3 lb or bigger chunk gets cut away from a chunk weighing at least 0.3 lbs the capertiller can survive as two capertillers. If it's cut across the width not down the length, it is in segments and would not like to have those sliced in half. Well if it only weighs 0.6lbs there's only one place on its whole length that can survive that split. But if it keeps on getting longer and longer there's more and more of it that could get chopped off and survive, and it can choose to let go of a part that is itself long enough to be split up and survive the process. Now there's reason to get 4 times as big as their 9 inch version. And then a bit more to grow more margin for error. And then, oops, the human is all full.
Only the one silly rule, the split up Duplication rule, and no need for argues about IQ or Will.
And since I'm already tweaking Possession in several ways, mostly to give the host IQ and add the 'tiller Will and make it far too powerful, then getting rid of silly rules is of the good. They'd have to buy up extra Duplicates to go with their Size Modifier, or buy them as extra lives, or SOME thing.
If someone is trying to surgically remove a capertiller, the 'tiller can run away, lose a chunk to go hide, and let most of itself get cut out. It would be a big sacrifice, but if someone missed the smaller chunk it could come back and have another go at possessing that same host, possibly while they were weak from the surgery and not prepared to resist them. Sneaky.
... and exactly like Stargate. I spend much time re-inventing the wheel.
Okay, I'm boring even myself now.
But the creatures are not boring at all.
Giant Capertillers are huuuuuuge and weigh at least 5 lbs per component mind with another 500lbs bit needed to connect them all together. A Green Anaconda that was 17 feet long was only 215 lbs. So, Giant Capertillers are bigger than the biggest snakes. And that's just to plug the first person in. It just keeps being longer to connect however many people want to plug in. It is also not smooth or scaled like a snake, it is soft and squashy and a bit wet and has pointy bits that in a parasitic capertiller wrap around the spine but in a giantm can wrap around a whole person.
Capertillers are tiny, and can fit inside large squirrels, let alone ordinary size people. They grow bigger too, but only so they can split into two pieces and relax sometimes, or hide in a dropped portion. They can squish themselves very small and are quite difficult to find in a large being.
Giant Capertillers are hard to miss, since the smallest survivable unit is 550lbs of squirmy thing like a big flat wide snake three times as long as a tall human. But just like their tinier cousins, as long as they have at least that much mass, they can survive splitting, though they may have to drop a few component minds to do so.
It is probably a good thing that they're optimised for aquatic environments, and can only lump along kind of like a seal on land. You can almost certainly outrun them.
If you see them coming.
But I've been thinking about all the rules tweaks I chose for my Giant Capertiller Gestalt Matrix Computer.
To start with, in the usual rules for a gestalt, if you plug one brain it, it becomes a complexity 9 computer capable of running 2 complexity 9 programs, and a mind emulation is complexity 9. So basically you can run (a) yourself and (b) any actual computing you want to do, both in the same body. That makes logical sense.
But you plug 8 more bodies in and you can still only run one mind in them, which doesn't make sense.
You get a lot more work out of them if you pretend they are separate computers that talk to each other.
But to do that you'd need 500lbs of computer each instead of just 50lbs of connector.
... I haven't decided how much the Giant Capertiller weighs. If it weighs 550lbs per mind, it could do that. But that's like four times as much as the humans it is plugged in to. Heavy.
Plus you wouldn't get one shared VR that way, you'd get many that talk to each other. I'd have to decide on data rates and suchlike.
It's unwieldy, complex, and stupid. I will not do that.
So you can have 9 bodies plugged in and still only run one mind. One mind of average intelligence. It's daft.
So I like the vatbrain biocomputer rules, because then you can run extra minds or very smart minds from the first time you plug in. 20 programs worth of them, in fact.
But when I plug in a regular size capertiller to be a backup brain, in some ways, it makes a lot of sense to call it an AI running on a single brain gestalt computer, because then you can limit its IQ relative to the complexity of the system, and know why it can't run without plugging in. It's a hard drive waiting for a computer. I knew that. Plus a gestalt component mind is unconscious while it works, so built in to that is the idea that the host is unaware while the capertiller is thinking. But with capertillers, I want two and only two minds able to work at any one time. And one of those needs to be unconscious if the capertiller wins. And I don't want to mess around counting how many programs a smart person takes up. And, fairly importantly, it does not weigh 550lbs. Quite a lot less really. But it is also two tech levels more advanced than the gestalt matrix. Advance a tech level, get smarter out of smaller. The tiny ones are the advanced models of the big ones. So I'm trying to figure if the ways it doesn't act like a digital mind outweigh the ways it does. There's no VR in there, it can't create its own world, it can't run programs, and it was never intended to interface with other technology. That's quite a list of not-digital. The only way it is digital is it depends on the complexity of the supporting system to have any IQ at all. For an AI, that's a simple taboo trait, and costs 0. So I just need to make the rule 'IQ cannot exceed that of the host' and stop worrying about making it match the Giant version. It's a different device of a different tech level that did away with many features in the quest to miniaturise. Old mobile phones did not do many of the things a mainframe did, even using the same variety of tech.
So it's not a digital mind, it's just a mind with a taboo trait limiting its intelligence to that of the host. Skills go with it, so it can get very good at a lot of things. A backup brain takes a number of months equal to the hosts IQ to create a full record of the host, one that can just go and run around being the host's brain if it feels the need. A possessing entity does not get that kind of access to a host, so it isn't Possession advantage that does that. And it isn't Extra Life (Copy), at least not for the symbiote. If it doesn't get conscious access to all that data then it's just a special effect on Racial Memory (Passive), and if it does, that's Racial Memory (Active). But the symbiote has to remain in a body for long enough to soak all that stuff up before it becomes part of its own memory. Before that they would have to let go of their host, let them become conscious, and either ask them or use the struggle for possession to wring data out of them. And every such attempt would give the host a chance of regaining control, making it immune to that particular 'tiller. But if the host willingly relinquishes control back to its 'tiller they could, like tok'ra, switch off, unless the host body immune system reacted unpleasantly (their HT exceeded their will and physically rejected the symbiote) or the 'tiller totally bungled the catch (crit/fail its will check for possession). I don't know what a fail would represent in that situation, sticking their neural probe legs in the wrong nerves? The possession equivalent of forgetting your password? Still, it's pretty unlikely once you buy your will up high enough.
But I've been calculating the will for initial possession purposes as if they already had an IQ, the IQ of the host body. They don't get access to that unless they win, in logic world. Hrm, it's like when a virus program loads on a computer it shouldn't be in, it runs as fast as the system allows, right up until the virus checker kicks it out. Similarly, grabbing all those nerves gives the 'tiller access to the hardware right up until the software or the immunities kick it out. There's sufficient logic there.
But if they do that there's more chance to possess a high IQ host, since there's less chance of rolling a fail, than there is of possessing an animal, because the 'tiller in an animal isn't bright enough to figure out what it's doing and might just bungle the plug in. That... is not a bug. They were designed to back up intelligent beings, not to pilot animals. And we wants them to try and borrow people who will know enough to not like it. Drama. That works out.
Okay, so the remaining constraint is, how long does it take for a blending to transfer data?
And I don't know, really. It shouldn't be instant, but it can't take a number of months equal to cumulative IQ of previous hosts, which would be the logical extension of the Backup Brain rule. Perhaps I shall just say that when two 'tillers like each other very much they go back to the sea and Stuff Happens.
A capertiller can split in half, and retain lots of data and some of its will and fatigue. Data is really easy to grow space for.
Data Storage: Additional built-in data storage can be purchased for $1 and 0.001 lb. per additional TB (at TL9). Multiply storage by 1,000 per TL after TL9.
Capertillers do things that are only possible at TL11, like being backup brains, but their data storage could be a lower tech level relic. They're also biogadget, like a vatbrain biocomputer, needing to weigh twice as much because of life support.
So each new bit of data storage weighs 0.002lb. If it's TL9 that's 1TB, TL10 1,000 TB, TL11 1,000,000 TB.
It is not TL11 because that's ridiculous.
One human mind backs up onto 100 TB.
Even at TL9 that's 0.2 lbs of data goop. You could store three minds in a can of soup. If their data storage is more advanced than that... you're just not going to run out of TB. Ever. At all.
In ten thousand years though?
2lb of capertiller is 1,000 TB at TL9. 10 minds. That's only one every thousand years.
But if the data goop is 2lbs of TL10, that's 1,000,000 TB or 10,000 minds, one a year. You wouldn't change hosts once a year. You wouldn't absorb them all in less than ten months unless they were particularly stupid.
So the capertiller has TL10 data storage and needs rather a lot less than 2lbs of goop.
If you say there's one new mind every twenty years for ten thousand years, continuous remembered experience since the cities were lost, that's 500 minds, or 50,000 TB. At TL10 that weighs only 0.1 lbs and we're back down to six to a soup can.
That's just the data storage parts of a capertiller. It still needs a body around them. So the capertiller might even fill the whole soup can.
Which is handy, because, as a parasite, it needs to fit inside a host body, even while remembering lots of things.
I know, I'm comparing weights not density or other useful dimensions. And the GURPS rules just give you size modifiers based on length, they say nothing about weight.
Dwarfism says about weight: Size Modifier -1. Choose your weight from the first line of the Build Table and reduce it by 15%.
The Build Table first line is for height range 4’4”-5’2” and skinny weight 40-80 lbs.
But on the Size Modifier tables 1.5 yards (4.5’) is SM-1, and that's 4' 6". So there's a bit of overlap.
Parasites have to be one SM smaller than the thing they're trying to possess. Which would mean a capertiller can be up to 4' 6" and still curl up inside a regular human.
... eeew?
But it's totally the wrong shape and sort of thing to weigh as much as a dwarf human, even a skinny dwarf, even if it was that long.
The spinal cord weighs approximately 35 g. I googled it. I had not known that before.
The spinal cord is long and skinny and made of nerves, like the remembering bits of a capertiller.
It's also where the capertiller plugs in.
That probably shouldn't work, but then, neither should goopy parasitic backup brains.
The spinal cord is only 45 cm long and then turns into a horse tail of nerves that run down the rest of the bony bit.
I am learning many things today.
Being longer than the thing it plugs in to would do the capertiller no extra good.
Half a yard long is SM-4. Capertillers can be SM-4 and lie along your backbone sticking their legs through the gaps. Or just grab the spinal nerves where they stick out.
I had made a formula where Will is related to Size Modifier, but I don't think I want to make their size as variable as my maths was. So I'll redo that idea.
The sticky outy bits of the back bones are all on the posterior side, so the capertiller stays comfy on the smooth anterior side.
I am learning words and making stuff up.
And yes, most of this is because I look at a goa'uld and just wonder where it all goes. I mean they're huge great squirmy things. There's a scan that shows one wrapped around the bones in the neck, Stargate Wiki page on goa'uld shows it. But there's much more squirmy thing when they're on the outside. And necks are quite busy. Can you really fit an invisible thingy in there? Plus, sticking a bit snake head into your brain just seems more likely to create disruption and permanent damage than possession.
... I'm criticising the plausibility of Stargate. *facepalm*
But I just want to put my parasites somewhere more roomy, with plenty plug in points. They don't need to be in the brain to access the brain. Eyes aren't.
So: 45cm long, weighing at least 0.1 lbs or or 45g, probably more so it can have life support. A Backup Brain is a Tiny Computer, which weighs 0.05 lbs, or 0.1 lbs as biocomputer. That doesn't seem sophisticated enough since even at TL11 it's only Complexity 6, but I don't want it to run a mind emulation, just remember them. It only needs much more basic responses to jump in a host. I set its IQ as 2 the same as a rattlesnake, and you can run an IQ2 volitional AI program (IQ/2)+3= Complexity 4 -1 for biocomputer, so IQ 2 VAI C3 program, in even a TL9 Complexity 3 tiny comp. And it's likely to be TL10, and still smarter. So, more weights and measures:
up to 45cm long, weighing at least 0.2 lbs or 90g. That's only like three times as much as a spinal cord. I probably need to add some weight to be all the little plug in arms. And it needs a power supply, but it gets that by leeching off the host body. It's just a feature of being a biogadget and doesn't have a weight multiplier attached, that's already taken care of by being a biocomputer. So, just the tiny arms, but those are to carry neural interface systems, which have no weight listed. It's not going to weigh more than 0.3 lbs unless I tell it to, basically. 135g, waaaay less than a can of soup. But if it was going to split in half each half would still need to be 0.3 lb so a growing 'tiller needs to be up to 0.6 lb or 300g before it can survive splitting with all its memories intact. If it gets split when its smaller than that it probably loses a lot of data storage. Amnesia time. But if the chunks of an injured 'tiller weigh more than 0.3lb they are in fact two injured capertillers. It is very goopy and can stretch and squish a lot. Its length isn't the important measurement, just its weight... which is unuseful when GURPS size modifiers are all in length and ignore weight.
An SM-4 humanoid, if it weighs 15% less per size modifier, is 18" tall and weighs 20 lbs. Searches say babies don't weigh that much at that size but they can't move around on their own neither. But snakes are more like the right model for a limbless capertiller. A snake that is sometimes 18 inches long can weigh 0.5 lbs or less, though there's loads of kinds of snakes and weights so that's difficult to pull relevant numbers. But the 'tiller weighs like a snake. That's about right then. Though it doesn't have bones or scales, so it should be on the light side.
A capertiller can be up to 18 inches long, and weighs around 0.3 lb, but probably more because it's trying to grow a second one. As soon as it can split up and both halves weigh 0.3 lb then it can survive getting cut in half. It is sneaky like that. It fits along the host body's spine so in a human it has plenty of room, and in most things SM-3 or bigger it doesn't even have to squish itself up much. To fit in a squirrel is trickier.
There's obviously a point past which it's just ridiculous to try and fit. The GURPS rules figure that out by size modifier, but I have instead weight. If one that is ready to split can still be SM-4 then one half that size can get as short as SM-5 and be perfectly healthy. That's like 9 inches long. I have found sorts of squirrels that are that size, 7 to 10 inch SM-5, like the regular red squirrels around here if you don't count their tails, but also there are big ones that are more than 10 inches long, so SM-4. Yes, that's a very big squirrel: an Indian Giant Squirrel, for instance, or Russet Ground Squirrel, also called the Large Toothed Souslik, or the Anatolian Souslik; a zizel is a souslik. But a regular grey squirrel can get up to that big, says wiki. A capertiller being SM-6 could fit into a sort of a squirrel that was SM-4, under the parasitic possession rules. Though it would make it a rather fat squirrel. A single capertiller can possess a squirrel, for sufficiently large values of squirrel.
The only reason they'd deliberately get big enough a human host might not be big enough is if their size lets them grow something useful, like more IQ of their very own, or higher will to possess bigger things. Computers don't grow a complexity until they weigh ten times more. To go from SM-5 9 inch long creature to ten times that makes them 2.5 yards long and SM+1. Humans are 0. They would outgrow a human long before they got any IQ benefit from their increased size. Is sad.
They could try possessing an elephant. There's a lot of room to grow in an elephant.
... Maybe it is best if there are no elephants in this setting.
... or, possibly, you end up with very smart possessed elephants. Huh.
Wait, I might be reading size modifiers wrong. You're supposed to round up, so it's things up to 7 inch or up to 10 inch. Oooh, maths. But then 18 inch is SM-4 but 9 inch is only SM-5. But then all those squirrels who are more than 10 inches long are SM-4. So they can still fit in a squirrel.
I will go up and fix it. Okay, is fixed now, only this paragraph stands as testament to my oops.
If it is not a computer it can get smarter as it gets longer. Then there would be actual point to getting longer, instead of just splitting up. I'd be making a new rule though. I think I'd rather stick to Will boosts rather than actual IQ. Big old capertillers can hold on in more places and that makes them better at possessing stuff. As long as the stuff is big enough. In small dumb things the host can't fight back mentally very well, but the capertiller can't borrow enough brain power to figure out where to plug in, so it's quite likely to fail at possession. In smart things the host can fight back mentally with a very high chance of success, but the capertiller is now borrowing smarts enough to keep up, and has a much higher chance of not just screwing up the attempt.
Oh, I've thought of a reason to get big: if a 0.3 lb or bigger chunk gets cut away from a chunk weighing at least 0.3 lbs the capertiller can survive as two capertillers. If it's cut across the width not down the length, it is in segments and would not like to have those sliced in half. Well if it only weighs 0.6lbs there's only one place on its whole length that can survive that split. But if it keeps on getting longer and longer there's more and more of it that could get chopped off and survive, and it can choose to let go of a part that is itself long enough to be split up and survive the process. Now there's reason to get 4 times as big as their 9 inch version. And then a bit more to grow more margin for error. And then, oops, the human is all full.
Only the one silly rule, the split up Duplication rule, and no need for argues about IQ or Will.
And since I'm already tweaking Possession in several ways, mostly to give the host IQ and add the 'tiller Will and make it far too powerful, then getting rid of silly rules is of the good. They'd have to buy up extra Duplicates to go with their Size Modifier, or buy them as extra lives, or SOME thing.
If someone is trying to surgically remove a capertiller, the 'tiller can run away, lose a chunk to go hide, and let most of itself get cut out. It would be a big sacrifice, but if someone missed the smaller chunk it could come back and have another go at possessing that same host, possibly while they were weak from the surgery and not prepared to resist them. Sneaky.
... and exactly like Stargate. I spend much time re-inventing the wheel.
Okay, I'm boring even myself now.
But the creatures are not boring at all.
Giant Capertillers are huuuuuuge and weigh at least 5 lbs per component mind with another 500lbs bit needed to connect them all together. A Green Anaconda that was 17 feet long was only 215 lbs. So, Giant Capertillers are bigger than the biggest snakes. And that's just to plug the first person in. It just keeps being longer to connect however many people want to plug in. It is also not smooth or scaled like a snake, it is soft and squashy and a bit wet and has pointy bits that in a parasitic capertiller wrap around the spine but in a giantm can wrap around a whole person.
Capertillers are tiny, and can fit inside large squirrels, let alone ordinary size people. They grow bigger too, but only so they can split into two pieces and relax sometimes, or hide in a dropped portion. They can squish themselves very small and are quite difficult to find in a large being.
Giant Capertillers are hard to miss, since the smallest survivable unit is 550lbs of squirmy thing like a big flat wide snake three times as long as a tall human. But just like their tinier cousins, as long as they have at least that much mass, they can survive splitting, though they may have to drop a few component minds to do so.
It is probably a good thing that they're optimised for aquatic environments, and can only lump along kind of like a seal on land. You can almost certainly outrun them.
If you see them coming.