Other Horizon Technologies in EP

14 posts / 0 new
Last post
ringringlingling ringringlingling's picture
Other Horizon Technologies in EP
Are there any other future or near future technologies that might revolutionize the EP setting? Something that would introduce profound philosophical questions or new insights into the human condition?
Xagroth Xagroth's picture
You mean aside from the
You mean aside from the digitalization of the human mind and the capabilities to re-program it? Or how the Deep Learning capabilities exhibited in Google Alpha Go (the second version, 4 days to win!) mean it's easier for us to make AGIs capable of really fast learning? Probably Drone Warfare will force a change in the game, since right now it's of so much relevance and the way it has been analyzed on the forum places it as the best way of deploying firepower (if not very discreet), plus the space combat analysis are also in a similar vein (mothership is little else than a cargo vessel with engines, deploys "weapon satellites" and "sensor satellites" and tries to neuter the enemy enegines). Google's assault on Big Data, with so many "easy to use" BI suites, coupled with the somewhat reduced population after the fall would also mean a shift in how the scanners work in Firewall. The Chinese have tested something we are not sure if it's teleportation or time travel, with data arriving at destination some time before it was sent... Also, the advances on data storage and miniaturization of data processors (there was some news about a new microprocessor by Intel with an expansion for a GPU that would act like support CPU/graphics card...) might slightly realign other details, but I doubt we'll see much aside from a little more understanding of the "computronium" that is Iapetus. Couple that with Tesla's advancements on batteries, and you might suddenly get a lot more egos per cubic meter than what EP considered a few years back. For example, the Sweet Dreams incident might suddenly go form 120.000 or so egos to 360.000 or even more... Also, 3D printing's rise might point to a change in paradigm to an extent Sunward-wise regarding fabricator machines. People is nowadays quite happy experimenting and desgining their own stuff for 3D printers (and a lot of them are really affordable), so it might be possible that win AF10 a lot of people design and print their own stuff instead of going for the corp-DRM locked. Of course, most of what people would self-design will be "open for all", but basic and for poor people. Another trend is the wearables... so instead of carrying stuff on bags and the like, the gizmos might be directly "sewn" on the clothes... much much more aggresively!
o11o1 o11o1's picture
The rise of good wearables is
The rise of good wearables is also a bit of a knock against implants. Why implant something when a better one is sure to come out in two years?
A slight smell of ions....
Xagroth Xagroth's picture
o11o1 wrote:The rise of good
o11o1 wrote:
The rise of good wearables is also a bit of a knock against implants. Why implant something when a better one is sure to come out in two years?
It depends on the type of wearable. Ectos are a wearable, but provide nothing more that mesh inserts, with the added point of "take them out, put them again, charge it, store it"... Granted, in this specific case most people want the disposability of the ectos. The main use I see for wearables is support stuff for what a Sentinel would have, or cheap alternatives to "paupers" who have a baseline morph and can't go to have modifications (because of money, beliefs, contractual obligations...). Of course, not having time to get the morph modified because of mission constraints is another reason to go with wearables (camo smartclothes come to mind here). For sentinels, disposable support and supplies would be the norm, kinda like military aircraft carrying external fuel tanks. The obvious stuff are Feedstock pre-packed for some selected blueprints, extra batteries/ammo, surveillance systems, comms enhancers/maskers, extra processing power, extra cold storage, etc... Another thing that is making some (quiet) headlines today and might be really relevant for EP are the "anti-cam stickers", adhesive printed patterns designed to confuse automated image recognition software, something that has been tested against google's systems (a 3d printed plastic turtle, and then another with two smal holes on the base and a different pattern of color and shell shapes lead to the system recognizing the first as a turtle, and the second as an assault rifle). While I doubt AGI or Uplifts will get so confused (there can be camo patterns specific against them, though, the same way there are some specific against human minds nowadays), the automated systems of the Panopticon would be another matter entirely, and while it will be possible to counter those stickers, new ones will be designed. And those are, in a way, visual basilisk hacks of the "you are not seeing what you see" kind.
MrWigggles MrWigggles's picture
AGI and Uplifts wouldnt be
AGI and Uplifts wouldnt be confused because they dont do machine learning pattern recongization, they're sapient self aware creatures.
CordialUltimate2 CordialUltimate2's picture
MrWigggles wrote:AGI and
MrWigggles wrote:
AGI and Uplifts wouldnt be confused because they dont do machine learning pattern recongization, they're sapient self aware creatures.
Image recognition in year 2017 is usually done by simulated neural networks which are trained for the task by machine learning algorithms. It is possible to do image recognition with hardcoded programs made "by hand" by programmer. But it is much easier, comparatively, to use neural networks for image recognition. Image recognition in human brain is conducted in analog neural network. Sufficiently advanced neural network could be indistinguishable in accuracy to human visual cortex. What you are missing though is processing power. Simulating neural networks is pretty resource intensive. So image recognition soft is optimised for efficiency. On the other hand biobrains are optimised for accuracy (correctness is essential to survival). And also they usually further process the image beyond first identification, which allows them to correct mistakes and identify discrepancies. Just ask yourself how many times you mistaken for a bush only to see that it is a TITANspawned fractal when it moved. Soooo that exposes a very big flaw in the logic of EP setting. With Transhumanity mastery of neural networks (clearly evident with widespread use of cyberbrains and psychosurgery) the limiting factor on any intelligence is processing power not the know-how and design of the intelligence itself. If we assume that much of the tasks performed by AGIs out indentures is dull and grueling then they clearly don't need ALL their "brainpower" to run the task. Therefore it is massively inefficient to use fully fledged egos where artfully pruned forks or specialty intelligence would suffice. It casts new light at the indenture and hypercapitalism making it seem almost benevolent where they accept massive cost* increases in exchange for reinstatement of ego for society as a whole. If we assume that "hypercorps" want to instantiate as many egos as possible it somewhat rationalises their underhand handling of nascent and rising corps. They have to eliminate newcomers who could run without inefficient indenture system and destabilise the instantiation operation. *This interpretation assumed that costs of running a full ego vs pruned forks are sufficiently different. EDIT: I also assume that intelligence is a continuous spectrum spanning from a toaster to a sapient being with very smooth and full curve/dataline for a biobeings between bacterium, single cellular organisms, protozoa, bugs, rodents, dolphins, humans etc. with AUTOMATED machines as sparse datapoints below the bacteria and all the rest of the way. In Eclipse Phase with titans and their creations extending way beyond human range. Also IMO intelligence doesn't require self-awareness.
Exurgents wanna eat your ass and you are low on ammo? Register to mobile gear catalogue at [url=http://eldrich.host]eldrich.host.mesh[/url]! ORDER NOW! FOR FREE PLASMA MINIMISSILE PACK! *explosive delivery options included
CordialUltimate2 CordialUltimate2's picture
Moderator please move this to
Moderator please move this to general discussion.
Exurgents wanna eat your ass and you are low on ammo? Register to mobile gear catalogue at [url=http://eldrich.host]eldrich.host.mesh[/url]! ORDER NOW! FOR FREE PLASMA MINIMISSILE PACK! *explosive delivery options included
Xagroth Xagroth's picture
Therefore it is massively inefficient to use fully fledged egos
Ok, first of all (before I forget): "camo" patterns do exist since the dawn of time, the difference with the stickers I mention would be essentially that they work against programs designed to recognize physical stuff by using cameras; while it becomes possible to "patch" this vulnerability on programs (or even digitized egos... but Psychosurgery would be more a "reprogram parts of an ego without crashing it... too badly", instead of "psychological treatment on steroids, IN SPACE!"), for a regular human it would need some neural rewiring (thus, evolution or genetics manipulation). Anyway, one could say that BIDI codes and this "camo stickers" are the "cuneiform level" of what Basilisk hacks are in the setting, remember that the Firewall book mentions how some strains have been "patched" (all of them from the time of the Fall, however) after some huge brute force testing. Remember, even nowadays some machines are "programmed" using barcodes! On the reason for indentured use in a Hypercapitalistic society: there are several things going on here, and in fact the "indentured contracts" by themselves are good and logical for that kind of society. First: Transhumanity is small, much of the population killed during the Fall, and the only attempt to produce a new generation of post-fall transhumans giving the Lost Generation. This means the hypercorps have a small base of consumers (so not as many people that buy stuff). This means low demand of goods and services. Second: even Sunward, the society is a post-scarcity one. However, we must remember the Offer-Demand graph: too much offer, too many demand, and you have an economical crisis where the value of money plummets down because the goods are overabundant. Third: Waste not, want not. It's faster to grab an ego willing to work his ass off for some time in order to get a body, than to make a psychosurgery operation to edit and prepare forks en masse, and copying forks has been forbidden by law precisely because #1... those "slave egos" earn no money, so they cannot spend money... so whatever they produce is superfluous for a capitalistic society, since no one has the money to pay for it, aside from the corps, leading us to #1... I think it's safer to assume that those thinking among this lines live in their own private habitat with the slaves they need. Fourth: Societal situations. Anything "AI-related" (AGIs, Infomorphs, synthmorphs...) is seen on a bad light in a lot of places, because of the Fall; as alien as anything humanity can produce, at least. This means that you cannot use AGIs, but LAIs (Limited AIs), which need a supervisor, which means having people herding systems at all times... Fifth: a consequence of #2, and why sleeves for indentures are as uncomfortable as possible without being detrimental for the job: motivation ("get out of this hellhole fast") and scarcity mindset ("the synthmorphs are inferior, I spent years sleeved on one", and the biomorphs take months to assemble), plus a synth with medichines requires less "maintenance" than a splicer with medichines (which still needs to sleep, eat, wash...). So the hypercorps solution is simple: generate future clients by way of "training" (indenture contract!) them with usefull skills for the future, mantaining the societal sctructure the infugee is used to (and expects), while mantaining a carefully balanced offer amount by way of different means and ways. Consider that with EP's tech it would be possible to instanciate all the infugees in cold storage in less than a month, by way of using nanoswarms to assemble steel morphs from some asteroids, and you can reposition other asteroids closer to the Sun for the production of energy... It's all about keeping the money circulating, in the end. Which is good and well, until some "bright corp" decides to run the scam of hiring the same ego, from backup, over and over... which provides little to no gain to the corp in both the short and long term: they have to train that ego over and over, give it a synthmorph to do the job, and they save, what, giving them a biomorph at the end of the contract? The other corps would quickly apply pressure over that corp to release a new consumer onto the herd ASAP. Which makes the fiction text on the venusian miner indentured who is programmed to kill himself (and wrecking the VERY expensive morph he uses!) not really about how hypercorps abuse clauses to keep you indentured forever, but something else...
ThatWhichNeverWas ThatWhichNeverWas's picture
Yeah, this should be on the General board.
-- Regarding the OP; As might be apparent, going forward from EP tech is pretty hard for us to imagine, if only because the full effect of the stuff they already have is still unfolding. Next-Gen for EP really begins hitting the SpaceMagic border: Wormholes, Kugelblitz Black Holes, energy to matter conversion, Picotech that sort of stuff. -- Regarding Wearables; Really, the biggest hit against implants in general was the creation of WiFi, since it allows the interface to be seperated from the device. That said, there are more than enough technologies which simply work better as an implant, such as medical monitoring and physical enhancement systems. Honestly, one thing that bugs me in EP is that Mesh Inserts get so little attention - a functional direct nervous system interface would cause societal change on the same level as the invention of the computer. It's one of the reasons I keep pushing for more Software as gear. -- Regarding Vision and Intelligence; As I understand it, the primary difference between AIs and Transhuman Intelligences is the ability to apply context and implicit information - an AI could indentify images as well as (if not better than) an Ego, but might not realise that (for example) something is suspiciously out of place - such as "why is that cardboard box moving?". The link between processing power and intelligence is there but it has some foibles, the simplest of which being Code Efficiency. Beyond that, you get an issue of diminishing returns with simple magnification; Imagine you had an AI or AGI intelligent as Einstein. You give it a complex problem, and it solves it in 1000 hours. You make it twice fast, and it solves it in 500 hours. Now you Reset, and run two instances of your AI on the same problem, and they solve it in 450 hours - they can divide up tasks to do things more efficently. You can keep doing this, but eventually you'll hit a plateau - you can't divide up tasks any better, and whilst you might have a thousand Einsteins running at once, they all approach things the same way and use the same ideas. Getting around this means introducing new Intelligences to the mix, using different perspectives and approaches to introduce new information to the process. It's noteworthy that it isn't [i]necessary[/i] for the new personalities to be as intelligent as AInstein (though it does help), they simply need approach the problem differently. If this seems familiar, it's because we do this today - it's called Crowdsolving, or more generally Collective Intelligence. EP just has the Dire-Half-Dragon version.
In the past we've had to compensate for weaknesses, finding quick solutions that only benefit a few. But what if we never need to feel weak or morally conflicted again?
sysop sysop's picture
(moving this over to General
(moving this over to General since this isn't about the playtest)
I fix broken things. If you need something fixed, mention it [url=/forums/suggestions/website-and-forum-suggestions]on the suggestions board[/url]. [color=red]I also sometimes speak as website administrator and/ moderator.[/color]
o11o1 o11o1's picture
So what -are- the most
So what -are- the most important societal effects of having ubiquitous mesh inserts? Functional telepathy with anyone closer than a light minute of you? Does this mean that you encounter all sorts of new views, or that the entire hab swiftly trends towards a social mono-culture? Externalized memories stored on whatever random bits of data storage you have in your house? If someone steals my tablet computer, does that mean they remove my ability to remember who was in my high school science class? If they faked that sort of thing, would I automatically notice or just chalk it up to being a memory I didn't worry much about? Seems like a clever spec Ops could insert themselves as a vague childhood acquaintance well enough to try "catching up over lunch" if people don't keep all their memories in their skull. Do Venus-Tech and Martian-Corp brand mesh inserts play nice with each other over internet? How often do I need to upgrade them and what happens if I try to switch manufacturers? Everytime I look at someone, facial recognition checks tell me their name, and if we share the same Rep Networks. At a glance, I can tell that my @-friends want me to shun that c-rep personality, and i can pull up a miles long list why that should be. The counter-arguments are stored on CivicNet, and effectively invisible to me if I choose not to look. If I can think of the above issues in five minutes, what counter measures have been taken for them, and how widely adopted are they in different polities?
A slight smell of ions....
Xagroth Xagroth's picture
o11o1 wrote:So what -are- the
o11o1 wrote:
So what -are- the most important societal effects of having ubiquitous mesh inserts? Functional telepathy with anyone closer than a light minute of you? Does this mean that you encounter all sorts of new views, or that the entire hab swiftly trends towards a social mono-culture?
I'd say it's more like being able to "/w [insert name of target here] [message]" in real life, kinda like in an MMO. Which is, in itself, an interesting way to live (you have the AR, so... I'm sure somebody is living in EP like inmersed in World of Warcraft-like graphics XD).
o11o1 wrote:
Externalized memories stored on whatever random bits of data storage you have in your house? If someone steals my tablet computer, does that mean they remove my ability to remember who was in my high school science class? If they faked that sort of thing, would I automatically notice or just chalk it up to being a memory I didn't worry much about? Seems like a clever spec Ops could insert themselves as a vague childhood acquaintance well enough to try "catching up over lunch" if people don't keep all their memories in their skull.
To be fair, the distributed stuff are processes, as in Ram/CPU. Storage hasn't been mentioned as spread upon devices, explicitly, and considering how much does a cortical stack stores, I'd say its not a real necessity. Cloud storage, however, with XP recordings... So if your tablet is stolen, your local mesh might lose some processing power and bandwith, but little else.
o11o1 wrote:
Do Venus-Tech and Martian-Corp brand mesh inserts play nice with each other over internet? How often do I need to upgrade them and what happens if I try to switch manufacturers?
This is not something the setting explains at all, but there are merits and demerits about compatibility. The first demerit about compatibility is the Exurgent Virus, we have had a Real Life example a month or so ago with the microprocessors vulnerability (Meltdown and Spectre) where all chips made in the past decade have the vulnerability... The first merit, of course, is the fast deployment of patches... Anyway, considering how things go, I'd say it's not really relevant. Mostly because even if the hardware is widely different, the software is the prime product in EP, so it's reasonable to think there will be emulation layers, firmware adapters and the like. As for interoperability, considering how there are no "brand variations" on implants or weapons, I'd waive it aswell, unless it's relevant for the running game (and only because it's interesting, important and does NOT become a chore), because this is not a game about shopping after all xD.
o11o1 wrote:
Everytime I look at someone, facial recognition checks tell me their name, and if we share the same Rep Networks. At a glance, I can tell that my @-friends want me to shun that c-rep personality, and i can pull up a miles long list why that should be. The counter-arguments are stored on CivicNet, and effectively invisible to me if I choose not to look. If I can think of the above issues in five minutes, what counter measures have been taken for them, and how widely adopted are they in different polities?
Well... yes and no. Facial recognition is essentially useless in EP: resleeving, implants, surgery... you need Kinesis software, which requires some degree of the skill, which also requires processing power, which is a chore unless you have a reason for it (call it the "detect evil" of a cleric/paladin...), the usual way to ID someone is a IFF tag that might include "links" to the sender's social media ("rep score account"), or at least the public ones, maybe even coded so only those of certain level in those circles can access the rep level (very likely for Guanxi, while for Eye Rep you will need to ask for the codename/"nick" on a secure channel). As for the "reasons to drop that guy", consider how your muse acts as your secretary, so your friends won't tell you "look at this page, he's a bad influence", it would be more like "here, take this data packet", your muse would peruse it, check it contains real data instead of forged data, and pass you her veredict... which you can check by yourself or (which is the norm) accept without a doubt.
ThatWhichNeverWas ThatWhichNeverWas's picture
Plug And Play Psychology.
--- [u]Upgrading.[/u] Whilst Mesh Inserts' processing and memory components might be subject upgrade, the sensors and filaments within the brain should rarely require replacement, and those are the only components which actually NEED to be within the brain – the rest can be located somewhere more convenient. Even then, Mesh Inserts are essentially a fancy I/O device, and shouldn't need upgrading any more often than modern keyboards, especially when EP storage and Distributed Processing is taken into account. The onboard computing is icing on the cake. --- [u]Telepathy.[/u] Generally communication is going to be text/speech as Xagroth says, but potentially with a "tone of voice" or emotional component to ease comprehension. In any case, as long as individuals have different preferences and experiences, you shouldn't get a monoculture; awareness of others' foible may foster acceptance, but could just as easily make differences more jarring. Most likely you'd get examples of both. --- [u]General Stuff.[/u] Whilst the common metaphor for Mesh Inserts or DNI use is to control things 'at the speed of thought', it's more accurate to say it allows input 'without thinking', at least if you can connect with nerves directly like Mesh inserts do. A quick example: Lift up your hand and look at the back, then look at your palm. Where did your intent to move change into the act of motion? For most if not all, there isn't a meaningful distinction between the two. You probably also haven't noticed your eyes moving from one word to the next as you read this. With Mesh inserts, this is the level of connection you have with [i]software[/i], creating a system where interacting with digital information becomes a nigh-instinctual act: there is no conceptual boundary between the intent to creating or searching for information and the act, and that information can be provided either consciously or at a subliminal level as appropriate. If you look at someone, you can see their MeshID in the same way you can see their hair colour, and other information can provided by faux-senses, so users are constantly aware of their (for example) GPS coordinates, current signal strength, blood sugar levels and heart rate, core temperature, neurohormone levels and brainwave activity on the same level as knowing their physical orientation. A fundamental consequence of this is that the creation of complex information like Art, scientific theories or product designs accelerates rapidly: even the most complex pieces of software can be used intuitively and all relevant information on techniques or equations is always on hand. Likewise, programs running on mesh inserts can feed their results directly into the senses, making them temporary additions to the User's awareness. I can give the reasoning behind this if desired.
In the past we've had to compensate for weaknesses, finding quick solutions that only benefit a few. But what if we never need to feel weak or morally conflicted again?
kigmatzomat kigmatzomat's picture
Nonvisual stimuli
IMO, one of the least considered features of mesh inserts enable a lot of information awareness via non-visual, or at least non text, stimuli. Tacnets could provide hot/cold/tingling/pain/crawling sensations on the "body" facing the threat, allowing the hind-brain reflexes to react immediately. Navigation software can make the right direction feel like "down hill" so you could find your way even with your eyes closed. Environmental hazards could have smells or the sensation of heat to provide awareness of proximity to danger. By the same token, the rep of people you meet could be represented in "auras", with color corresponding to network at brightness/size representing rating.
I'm not rules lawyer, I'm a rules engineer.