Welcome! These forums will be deactivated by the end of this year. The conversation continues in a new morph over on Discord! Please join us there for a more active conversation and the occasional opportunity to ask developers questions directly! Go to the PS+ Discord Server.

How does it feel to be tiny?

20 posts / 0 new
Last post
Jaberwo Jaberwo's picture
How does it feel to be tiny?
I am playing around with the thought of creating a character sleeved into a swarmanoid who constructed his own little habitat somwhere, filled with bonsai trees and similar downscaled furniture and equipment. But I know that a lot of things are different when you change the scale of things, like ants being able to move several times their own weight and surviving any fall because the impact simply doesn't damage them etc. Has anyone got some general guidelines to what changes and why or perhaps experience with playing a swarmanoid and interacting with the phyiscal world in other ways than interfacing with electronics?
The Green Slime The Green Slime's picture
Maybe the TVTropes for the
Maybe the TVTropes for the [url=http://tvtropes.org/pmwiki/pmwiki.php/Main/IncredibleShrinkingMan]Incred... Shrinking Man[/url] will be helpful (also [url=http://tvtropes.org/pmwiki/pmwiki.php/Film/HoneyIShrunkTheKids]Honey I Shrunk The Kids[/url] and [url=http://tvtropes.org/pmwiki/pmwiki.php/Series/LandOfTheGiants?from=Main.L... of the Giants[/url]). One of the commonest mistakes in depictions of life at the micro-scale is forgetting the surface tension of fluids, and the amplified effects of heat on smaller bodies.
Baalbamoth Baalbamoth's picture
Cant find it....
I remember reading about a habitat where all the synths were ant sized and the whole city/hab was built to scale, i dont for the life of me remember the name, any other fluff researchers remember?
"what do I want? The usual — hundreds of grandchildren, complete dominion over the known worlds, and the pleasure of hearing that all my enemies have died in highly improbable accidents that cannot be connected to me."
Jaberwo Jaberwo's picture
Thank you
Found it: p. 88 Sunward. Just what I was thinking about, but unfortunately not very detailed.
Baalbamoth Baalbamoth's picture
Got a hab name or link?
Got a hab name or link?
"what do I want? The usual — hundreds of grandchildren, complete dominion over the known worlds, and the pleasure of hearing that all my enemies have died in highly improbable accidents that cannot be connected to me."
Jaberwo Jaberwo's picture
Here you go
THE COLONY Allegiance: Lunar-Lagrange Alliance Primary Languages: Javanese The Colony is micro-sized, in the literal sense. All members of the settlement are sleeved in insect-sized (though not necessarily insectoid) morphs and the living environment is built to scale. Taking up only a small side cavern in the larger Karpola habitat, the Colony exhibits a transhuman lifestyle that still has all of the same amenities (nanofabrication, the mesh, resleeving, etc.) while being far less resource intensive. The Colony is a popular destination for Lunar tourists who wish to live small for a brief period. (It's really just called "The Colony")
Arenamontanus Arenamontanus's picture
Ah, this is fun! Scaling laws
Ah, this is fun! Scaling laws is one of the coolest ways of thinking about the world. Drexler sometimes drills me in this; check out http://e-drexler.com/d/06/00/Nanosystems/ch2/chapter2_2.html http://e-drexler.com/d/06/00/Nanosystems/ch2/chapter2_3.html http://e-drexler.com/d/06/00/Nanosystems/ch2/chapter2_7.html When you reduce your length by a factor x, your mass goes down by a factor of x^3 and your cross-section as x^2. This means that your muscles/actuators get weaker by a factor of x^2 (since strength depends on the cross-section area of your muscles), but your weight has decreased even more (x^3) - gravity will become much weaker, and you will feel stronger. Think of ants: lifting several times your body weight is no longer hard, and falling from a height much less dangerous. Your skeleton (whatever it is) is going to be relatively stiffer. The time it takes to move something like a limb goes down: its mass is x^3 but your muscles scale as x^2, so it can be accelerated like 1/x - if you are 1% of your original size you can now move a hundred times faster. Typical motions also speed up as 1/x - it takes shorter time to move a tiny arm. So your life speeds up, while big people seem to move slowly and ponderously. Fluids become more viscous from your perspective. Air has a perceptible drag (terminal velocity scales as sqrt(x) - as you become smaller you fall more slowly, and reach terminal velocity much earlier). Wind and convection currents (fire!) are much more powerful. Liquids are indeed surrounded by a heavy surface tension. Friction scales as x^2, so it is about as troublesome for big and small. However, Drexler notes that wear is worse for small things: stuff might wear out in time x (so if you can use your normal-sized morph for 10 years, at 1% size morphs might wear out in a month) You heat up or cool down much faster - you have a far bigger surface compared to your volume, so you lose or gain heat from the environment like 1/x. This is why small mammals and birds need so much food to stay alive: they can easily freeze to death in fairly warm temperatures. And don't get too close to a hot object, because you can fry equally easy. In fact, since heat conduction happens across shorter distances it is also faster, plus you have a smaller thermal capacity: thermal equilibrium times scale as x^2. Being small also means your eyes or other sensors will be smaller and hence get less energy - the total energy scales as x^2. So the world will be darker and quieter unless you get more sensitive eyes. And visual focus becomes tricky at a distance: much of the world will be like a distant landscape rather than in 3D. Low frequency sounds are harder to hear and make, ultrasound easier. In short, you become a fast tiny superman - strong, fast, hard to hurt, able to fly. Ob SF reference: the start of The Risen Empire by Scott Westerfeld. Micro-dogfighting!
Extropian
fafromnice fafromnice's picture
too much information !!!
too much information !!! being little will be incredibly hard if you're not in a controled environnement for the Colony, I was wondering : a micro size morph can't have a full cyber-brain it's probably too small so do you think the habitant's ego are in a server somewhere, a cloud like cyber brain ? it's seem that this problem can be found in the swarmanoid morph it seems to me to be a great security risk

What do you mean a butterfly cause this ? How a butterfly can cause an enviromental system overload on the other side of a 10 000 egos habitat ?

Quincey Forder Quincey Forder's picture
if you want to get a nice idea for tiny adventures...
...Check Epic from Blue Sky Studio There also is also some kind of time dilatation that make 1/1 scale individual seem so slow. Move slow, talk slow (Knob and Ronin has a great fun mocking the poor scientist about it) Though I don't know if the time difference is a fact, either real or perceived, but other stuff like the jump and stuff fit with your explanation, Arena
[center] Q U I N C E Y ^_*_^ F O R D E R [/center] Remember The Cant! [img]http://tinyurl.com/h8azy78[/img] [img]http://i249.photobucket.com/albums/gg205/tachistarfire/theeye_fanzine_us...
Arenamontanus Arenamontanus's picture
Small creatures do have a
Small creatures do have a fast world - signals in their bodies travel shorter distances, they can oscillate their limbs faster, they are stronger relative to their mass, and so on. Think about trying to catch small insects. It is not super-fast, but proportionally faster (things get amazing when you go to nanoscale). I think you can distribute cyberbrains across many swarmbots, but beyond some scale it is likely easier to have them in a server and just seamlessly telepresence. Security is indeed an issue (and swarmanoids might worry about somebody stealing part of their ego - they better use homeomorphic encryption).
Extropian
fafromnice fafromnice's picture
homeomorphic ?
homeomorphic ?
What do you mean a butterfly cause this ? How a butterfly can cause an enviromental system overload on the other side of a 10 000 egos habitat ?
Arenamontanus Arenamontanus's picture
fafromnice wrote:homeomorphic
fafromnice wrote:
homeomorphic ?
I meant Homomorphic encryption - sorry, I can't keep my homorphisms from my homeomorphisms. The idea is that you perform encrypted computations - the processor or somebody eavesdropping on it doesn't know what it is calculating, it only becomes visible when decrypted with the right key. This is how I would ensure my mental privacy.
Extropian
NewtonPulsifer NewtonPulsifer's picture
Homomorphic encryption is
Homomorphic encryption is very very very slow. So to suppose it is feasible in Eclipse Phase for real time networking might be a stretch. There certainly could be mathematics breakthroughs to get it faster than the 100k to 10billion times slower (than plaintext) it is now, but it probably still isn't fast enough.
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto
Baalbamoth Baalbamoth's picture
too much of this thread has not been about me.
Chris Hardwick (from the nerdist) spelled it out right in his HBO comedy special. I, gentlemen, am a nerd. I am not a geek and thus cannot understand the mathimatics and physics your using above. if you can... please give me some easy to follow broken down rules to implement the above physics... If I was running the tiny game... I'd probably just make some minor adjustments to surface tension, falling, etc. but I wouldent go so far as to make vision dim etc. I doubt I would have ever even considered it.
"what do I want? The usual — hundreds of grandchildren, complete dominion over the known worlds, and the pleasure of hearing that all my enemies have died in highly improbable accidents that cannot be connected to me."
Arenamontanus Arenamontanus's picture
NewtonPulsifer wrote
NewtonPulsifer wrote:
Homomorphic encryption is very very very slow. So to suppose it is feasible in Eclipse Phase for real time networking might be a stretch. There certainly could be mathematics breakthroughs to get it faster than the 100k to 10billion times slower (than plaintext) it is now, but it probably still isn't fast enough.
Why don't you think there could be such breakthroughs? Consider how making a discrete Fourier transform went from O(N^2) to O(N log(N)), or how multiplication of n digit numbers went from O(N^2) to O(N^1.585) to O(N^1.465) to O(N log(N) 2^(log* N)). Homomorphic or blind computation doesn't seem to be a problem that "has to" be in NP. However, I found this paper, which does indeed argue that homomorphic encryption might intrinsically require more computational power than public key systems. However, the quantum case looks much better - and in EP quantum computers are part of the canon. In many ways quantum computing is blind by its nature, since you cannot peek on the contents of the computation without making a detectable mess.
Extropian
Arenamontanus Arenamontanus's picture
By the way: "FYI: How Do
By the way: "FYI: How Do Mosquitoes Survive Rainstorms? " http://www.popsci.com/science/article/2013-06/fyi-how-do-mosquitoes-surv...
Extropian
NewtonPulsifer NewtonPulsifer's picture
Arenamontanus wrote
Arenamontanus wrote:
NewtonPulsifer wrote:
Homomorphic encryption is very very very slow. So to suppose it is feasible in Eclipse Phase for real time networking might be a stretch. There certainly could be mathematics breakthroughs to get it faster than the 100k to 10billion times slower (than plaintext) it is now, but it probably still isn't fast enough.
Why don't you think there could be such breakthroughs? Consider how making a discrete Fourier transform went from O(N^2) to O(N log(N)), or how multiplication of n digit numbers went from O(N^2) to O(N^1.585) to O(N^1.465) to O(N log(N) 2^(log* N)).
Going from computing at 10billion times slower (N=100,000, O(N^2)) to 16 million times slower (O(N log(N) 2^(log* N)) or 500,000 times slower(O(N^2) to O(N log(N)) means it is still infeasible for general computing and real-time networking. Saying 100billion to 100,000 times slower was already giving you the maximum potential jackpot (O(N^1) possible for 100,000 - going from 1x10^10 to 1x10^5.
Arenamontanus wrote:
Homomorphic or blind computation doesn't seem to be a problem that "has to" be in NP.
All homomorphic problems are at least NP-complete, so yes they have to be in NP. You cannot make a practical (fast) encryption scheme based on NP-complete or NP-hard problems.
Arenamontanus wrote:
However, I found this paper, which does indeed argue that homomorphic encryption might intrinsically require more computational power than public key systems.
Link broken? Here it is this one.
Arenamontanus wrote:
However, the quantum case looks much better - and in EP quantum computers are part of the canon. In many ways quantum computing is blind by its nature, since you cannot peek on the contents of the computation without making a detectable mess.
To my knowledge no one has ever shown that an NP-complete problem can be solved more quickly on a quantum computer than on a classical computer - the evidence so far actually points to it being impossible on adiabatic quantum computers - but since adiabatic quantum computers are polynomially equivalent to conventional quantum computation (paper link), its the same conclusion. Quantum blind computing will only work for BQP problems. [img]http://upload.wikimedia.org/wikipedia/commons/1/1d/BQP_complexity_class_... Quantum computers are great for a narrow set of computing problems only. In addtion, in EP canon they're way more expensive than cyberbrains or ecto/insert CPUs, and possibly much more massive and power hungry as well (book doesn't say how big they are that I know of). In addition, wiring up a swarmanoid with 1000 miniature radio farcasters would cost 250,000 credits. Ouch. I'd like to be clear I'm not saying homomorphic encryption (or blind computing - although similar they have fundamental differences) has no place. They're useful things for it to do even if it is 100,000 times slower or limited to a narrow set of computing problems (BQP). However its practical uses are limited due to that slowdown and/or narrowness. You could say use it as a "black box" that spits out encryption keys for you to use as master keys - but if a crypto key requires say 20 milliseconds to calculate/create without homomorphic encryption - the "black box" is going to take 2000 seconds to do the same thing. Also, the data still needs to be decrypted at some point by this master key so the data can be stolen at that time. EDIT:spelling
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto
NewtonPulsifer NewtonPulsifer's picture
To answer the OP question,
To answer the OP question, your big limitations are going to be energy storage and heat dissipation. At 1/10 scale person is 1/1000th the volume. So if you do use all your potential strength and speed, you'll use up your energy very very quickly compared to a normal sized person. So you need to either be constantly eating, or better, be a synth with a quick charge battery that is charged often. The alternative is to be much more efficient in your energy usage (slow and weak).
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto
Jaberwo Jaberwo's picture
I was under the impression
I was under the impression that the blind computing with quantum computers relied on the inherent nature of the quantum computer, not on it's ability to solve certain problems better (or it's non existent gigantic magic computing power) and that it was as safe as quantum encryption (the thing where you meassure and check if the deviation is too high). Isn't this fundamentally different from homeomorphic encryption schemes?
NewtonPulsifer NewtonPulsifer's picture
Jaberwo wrote:I was under the
Jaberwo wrote:
I was under the impression that the blind computing with quantum computers relied on the inherent nature of the quantum computer, not on it's ability to solve certain problems better (or it's non existent gigantic magic computing power) and that it was as safe as quantum encryption (the thing where you meassure and check if the deviation is too high). Isn't this fundamentally different from homeomorphic encryption schemes?
Not how I understand it. For instance no NP-hard function can be computed blindly - see "On Hiding Information from an Oracle (1989)" by by Martín Abadi , Joan Feigenbaum , Joe Kilian Here's a good paper good paper on arxiv "Universal blind quantum computation (2008)" and a pdf slideshow by one of the authors. Check out page 1 of the paper to see what you can do with it (its limited basically to the things quantum computers are good at - factoring, BQP problems, measuring/processing quantum states/inputs). Bonus if you're interested: Here's a link to a post by another of the authors of that paper explaining the difference between blind computing and homomorphic encryption
"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto