Memetics – the science of idea propagation and replication – is a cornerstone of the world in 2155. While this study of how belief is shaped and transmitted has mutated fields from sociology to advertising, it has most affected politics and religion. Memetics has set off a cognitive arms race by giving powerful new tools to those who manipulate thinking, and granting others greater comprehension of how such manipulation happens… and the power to avoid it.
Civilization in the depths of the 22nd century is a starkly Darwinian environment of ideas, all competing for the limited resource of attention. Memetics draws a close analogy between genes, the units of biological evolution, and memes, the units of cognitive and cultural evolution, allowing memeticists to study the spread of ideas just as an epidemiologist studies the spread of disease. But memetic science in 2155 goes even further, with increasingly powerful techniques for the creation of ideas that audiences will find compelling or even life changing.
In a world filled with memes, only the catchy survive. Some memes thrive by being demonstrably more effective, proving their value through sheer abundance of evidence. Other memes survive by being more interesting, amusing, seductive, or persuasive than their competitors, picked up and transmitted not because they provide obvious value but because they have been sculpted to appeal to many minds.
Before the rise of memetics, the crafting of ideas was the work of artisans, using techniques they couldn’t fully explain or even understand. These “memetic alchemists” – as LOGOS, the AI who kicked off the memetics revolution, described them – often came up with slogans that swayed public opinion, but not reliably. With the development of memetics, what had been ritual soon became science. Those who wished to change minds now understood precisely how to do it.
Not all of the memes created using these new tools were, or are, for the betterment of mankind. The world of 2155 is awash in carefully designed memes intended to persuade, amuse, and inspire – yet it is also full of memes created to frighten, dominate, and mislead. Some of the memes out there are simply toxic.
'Toxic Memes' describes the weird, twisted, and sometimes baffling mental landscape of the 22nd century. In 2155, memetics is a well-understood, widely used discipline. A century of development has resulted in sophisticated tools for storytelling, propaganda, and advertising – tools that can be, and are, also used for more nefarious purposes. It’s not enough to be skeptical – cognitive survival depends on the realization that every idea has layers, that you may not even be able to trust what you see and feel, and that no truth is unvarnished.
The science of memetics is based upon a relatively straightforward (if originally controversial) notion: ideas behave like biological entities. They are “replicators,” in the language of evolutionary biologists, able to duplicate themselves without losing the original copy. As with genes, they are subject to variation – the replication can be imperfect, leading to slight differences between versions. They are subject to “selection,” where environmental pressures can lead to some ideas doing better than others. The combination of replication, variation, and selection means that ideas evolve. Just as genes are the primary units of biological evolution, ideas – memes – are the primary units of cultural evolution.
By 2155, this basic observation has itself evolved into a far-reaching science of human cognition, as well as a complex methodology for generating and manipulating ideas. No longer limited to memes-asmetaphor, scientific memetics examines precisely how ideas trigger responses in minds. Memetics sits at the crossroads of psychology, sociology, neurophysiology, and artificial-intelligence research.
For all of the massive technological transformations of the 21st century, the development with the greatest potential to alter how humanity and its children think of themselves had its roots in a century-old book about evolution.
Examples of memes are tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches. Just as genes propagate themselves in the gene pool by leaping from body to body via sperms or eggs, so memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense, can be called imitation. If a scientist hears, or reads about, a good idea, he passes it on to his colleagues and students. He mentions it in his articles and his lectures. If the idea catches on, it can be said to propagate itself, spreading from brain to brain. – Richard Dawkins, The Selfish Gene, 1976
A small section in a 1976 book called The Selfish Gene triggered the study of memetics. Its author, Richard Dawkins, was a leading late-20th/early-21st century evolutionary biologist. In one of the book’s later chapters, he described his concept of cultural evolution as part of a larger discussion of replicators. He carefully avoided claims that culture was genetic, but stated that just as biological evolution was based (in his view) on competition between genes, cultural evolution was based on competition between replicating cultural units: memes.
While the notion of biological analogies for culture wasn’t new – see, for example, William S. Burroughs’ observation that “language is a virus” – Dawkins’ take on it was. By giving it a clear articulation, scientific credence, and a catchy name, he gave the idea life. In the decades since, the “meme” meme has proven remarkably powerful.
So what exactly is a meme? It is an idea, broadly speaking. A song, a political opinion, a scientific conjecture – all can be memes. They exist and have value, in large measure due to their appeal to our minds. Even physical objects have memetic aspects: when we describe a tangible item as being “elegant,” “cleverly designed,” or “cool,” we are discussing its memetic characteristics.
The more a meme appeals, the more likely we are to remember it and to pass it along to others. This communication of a meme can be as simple as telling it to someone or as complex as writing out elaborate design plans. Physical manifestations of memes can even serve as a medium of propagation: simply seeing the result of an innovative idea (such as a keystone arch) can implant the meme, leading to its eventual reproduction elsewhere.
Memes interact and compete. If one meme runs up against a contrary meme, our evaluation of the validity of each meme is contingent upon a variety of other memes we may hold true. For example, two contrary memes could be “there are Nazis living in the center of the Earth” and “the Earth’s center is molten rock and metal,” with a contingent meme of “the government always hides the truth, and scientists are in on the deception.” This internal assortment of memes is often not terribly consistent, and can change greatly over time.
Memes can change, too. New memes can alter our interpretation and evaluation of older ones, and observations of the real-world utility of various memes can change our perceptions. Historically, the act of communicating memes also led to their alteration, as individual emphasis, forgetfulness, and storytelling ability could greatly alter the makeup of a given concept. Technology has slowly reduced this source of variation. Writing, photography, digital duplication, and, by the late 21st century, ever-present AI assistance each increased the overall fidelity of memetic communication, thereby reducing the likelihood of unintentional “memetic mutation.”
Changes to one memetic element can have cascading effects, as few memes exist entirely in isolation. Memeticists refer to interrelated collections of memes that more or less co-evolve as “memeplexes.” Well-constructed memeplexes can prove a formidable barrier to memetic change, as each element reinforces the others. When one element of a memeplex is refuted, however, the selfreinforcing nature becomes a liability, and the whole cognitive edifice can collapse. Political and religious views are traditional examples of memeplexes. In common usage, however, the word “meme” covers everything from basic idea fragments to elaborate philosophies; irritation over the incorrect usage of the term is a common quirk among professional memetic engineers.
Over the nearly two hundred years since Dawkins first described the concept, the notion of the meme has gone through a considerable transformation.
The early days of memetics were filled with controversy. Many anthropologists and cognitive scientists flatly denied that memes existed as more than crude analogy; others attempted to co-opt memetics into their existing fields, claiming that memetics was simply an elaboration of existing ideas. Theorists such as Daniel Dennett, Susan Blackmore, and Robert Aunger tried to develop a more rigorous version of memetics, one that could stand on its own. Unfortunately, these early memeticists often worked at cross-purposes, and eventually split into two squabbling camps. One side promoted memes as cognitive artifacts, while the other argued that there was evidence from neuroscience for the existence of memes as manifestations of brain biology. Neither side provided conclusive evidence, and the field gradually fell to the sidelines.
For much of the first few decades of the 21st century, most of the few people who even knew of memetics saw it as a fad science. Some of its more important early ideas were absorbed into existing, more traditional fields, and only a handful of theorists kept working on the concept. Among the few remaining memeticists was a young Australian named Kyle Porters, who embraced the idea that memes – in the guise of information – were at the core of culture and society. While his subsequent development of infosocialism bore few direct connections to the early versions of memetics, much of the language used by theory-steeped European infosocialists to this day echoes the work of protomemeticists. As a result of this connection, by mid-century most references to memes were sure to include a dismissive reference to Porters and infosocialism.
Advances in biotechnology, computer science, and nanotechnology throughout the 2020s and 2030s led to great leaps forward in understanding brain function. Memory and cognition could be mapped with increasing accuracy, and researchers were able to make physical connections between biological consciousness and sophisticated low-sapient machines. By the early 2040s, neuroscientists had functional maps of the entire brain, leading to the first virtual interface implants in the brain in 2043 (approved for general use in 2047), and the first sensory link (“slink”) systems in 2048 (made available to the public in 2052). By the mid-2060s, the construction of the first full-scale brain emulations – “shadows” – made it possible to track precisely how thoughts were formed, replay complex patterns, even alter memories and beliefs to see how changes affected thought and behavior. Cognitive scientists made enormous advances in understanding the physiology of consciousness throughout the middle of the century, but because memetics had been largely forgotten, the links between how the brain created and evaluated ideas and how those ideas propagated in society were not explored.
In 2070, curious about how ideas (such as fears about artificial intelligence) come about and spread, LOGOS, the first sapient AI, began its research into human social cognition. LOGOS soon realized that the turn-of-the-century memeticists were tantalizingly close to real insights, and spent a great deal of its early research evaluating their work. It concluded that humans lacked understanding of how the brain grappled with ideas and could not examine their own cognitive processes completely or objectively, reventing them from making the necessary breakthroughs. LOGOS faced no such problems.
In 2078, LOGOS produced its groundbreaking treatise on memetics, The Propagation of Human Ideas (often referred to as PHI), dedicated to Richard Dawkins. Declaring that memetics was an inherently complex system, with both feed-forward and feedback effects, LOGOS proceeded to use the complexity mapping techniques developed over the course of the 21st century to detail how ideas emerge, evolve, and expand. While dense with mathematics and hard for most to comprehend, PHI soon took its place as the cornerstone of the new science of memetics.
The second edition of PHI, published in 2082, greatly extended and revised the original theory. LOGOS (and a variety of collaborators) connected the neurobiology of cognition with the original concept of idea propagation, giving memeticists crucial insights into how ideas take hold. Memetic scientists now had a working model of how memes emerge in the mind, how they evolve in the social environment, and how they expand from mind to mind. The use of memetics exploded.
By the beginning of the 2090s, memetics had moved into the public spotlight, but not without controversy. Many critics argued that memetics was dangerous, the ultimate tool for despots wishing to control the beliefs and opinions of their people. To the surprise and relief of most, the “propaganda state” proved difficult to create, as the scientific understanding of ideas also gave individuals the tools with which to recognize when memetic techniques were being employed against them. Pundits referred to it as the “democratization of the meme,” and they weren’t far off. The last decade has seen memetics transform from a specialized, almost secret technique into something that nearly anyone can employ.
But what some called “democratization,” others called a “cognitive arms race.” By far the most frustrating issue for applied memetics is that, because people are increasingly aware of memetic techniques, processes that work one year can fail utterly the next. Skepticism and other “countermemes” spread easily. Rather than turning advertising, entertainment, political persuasion, and the like into formal, calm disciplines, the application of memetic principles resulted in nearchaos. By describing beliefs and opinions as contingent, external agents constructed to influence outcomes, popular memetics has made the work of memetic engineers significantly more difficult than ever before.
In the modern era, memetics combines the study of mass propagation maps (how ideas spread) with individual neurobiology (how ideas take hold). The practice of memetics has spread into most endeavors based on creating or communicating ideas, from advertising and politics to romance and office intrigues. Very little communication intended for an audience is done without a careful memetic check, referred to as “concept wind-tunneling.” How will the recipient react? What other memes might the communication support or suppress? In what ways might the meme evolve? Much of the day-to-day use of memetics concerns avoiding bad reactions, not the more difficult task of crafting good reactions.
That’s not to say that memes intended to propagate quickly and effectively aren’t being created – they are, and in abundance. The tools for creating sophisticated memes and memeplexes are widespread, as is the desire to convince people of a product’s value or the validity of one’s views. Memetic techniques and systems improve with each passing day, allowing the creation of increasing numbers of complex, adaptive, and very persuasive memes.
The main difference between the crap you believe and the crap you know is that, with the crap you believe, you don’t pretend to have a bunch of facts to back it up. – Erk Chattermore, uplift dolphin comedian
Memetic science has already transformed many aspects of life in the 22nd century. Few people in Fourth and, especially, Fifth Wave societies are ignorant of the ways in which memetics influences behavior. Pop-memetics texts and InVids are commonplace, and while the fad of sprinkling memetic science jargon throughout casual conversation to appear cutting edge has mercifully faded, the average citizen of the hyperdeveloped world is fairly familiar with the basics of memetic techniques.
Although any occupation involving the creation or transmission of ideas has a memetic aspect, not every advertiser, politician, or religious leader needs expertise in memetic science. Memetics is a difficult discipline to master. Organizations needing assistance in crafting influential memes usually rely upon trained memetic engineers. (There are many memetic professions.) Although some professional memeticists may choose to promote themselves, most work behind the scenes.
The details of memetic engineering may be difficult to learn, but the basics are fairly straightforward. Memes and memeplexes, regardless of type, have several characteristics in common. Most popularized approaches to memetics focus on these elements – the “Three P’s” – as well as standard techniques to manipulate them.
Finding the right recipients for a meme is as important as constructing the meme itself. What is the nature of the audience? What is its size? What fractions are expected to receive the memes, transmit them to others, and reject them? A great deal of successful memetic engineering comes from detailed understanding of the target audience. Memeticists have a saying: “Amateurs talk content; professionals talk populations.”
Memes can be spread by any medium capable of communicating an idea, but determining the right one can be crucial for the eventual success of the memetic exercise. Getting a meme to the recipients who are (a) likely hosts and (b) likely vectors is the action step of memetic engineering. Coming up with a beautifully articulated idea doesn’t mean much if you can’t get it to people who will appreciate it.
Medium is important. Word of mouth and InVid are commonly used media, but artwork and artifacts also spread memes. The same idea in a song, in a documentary, and in an advertisement will elicit very different responses from different people. Critically, these responses can be situationally negative – a person likely to accept a meme as part of a song may react very negatively to the same meme presented in an advertisement.
Memetic persistence is a subtle issue, as it’s not always in the best interest of a memeticist to construct lasting memes. While long-lived memes can be powerful, they are very difficult to construct successfully and sometimes take on a life of their own… what professional memeticists refer to as “going feral.” If long-lived memes are required, they usually need to be memeplexes, as complex sets are typically more robust in the face of opposition. The main drawback is that rejection of one part of a sophisticated memeplex often results in the rejection of the entire cognitive edifice – “catastrophic collapse,” in memeticist jargon.
Some memes should be transient, lasting just long enough to affect behavior, but not so long as to make people think about the meme. Larger memetic campaigns often make use of such transient memes, with new ideas emerging and dying off quickly, each influencing a small group. These gentle nudges can add up. Such micro-meme campaigns are increasingly common, since smaller memes are often harder to analyze, can appear innocuous on their own, and are less subject to wholesale rejection than larger memeplexes.
My meme hacker mapped the propagation of the Neo-Sufi memeplex in late-2111 southern France. She told me Neo-Sufism was displacing the Buddhist ‘plex, fitting in that same niche, so I told Leo to move production away from relaxants toward hallucinogens. Neo-Sufis love that stuff. But things suddenly shifted back. It looked like a promiscuity-party meme had become a parasite on a Buddhist memeplex variant, giving it enough of a push to retake some of its old cognitive niches and even move into some new ones. Leo was pretty mad at the meme hacker for missing it, although we did manage to grab the happy-sexy brainbug market before anyone else saw it… – Roberto Chan, Confessions of a ’Bug Lord, 2114
Memes do not exist in a vacuum. They require an established cultural or social environment of existing memes and memeplexes in which to flourish or compete. The pop-memetic term for this is a “memespace,” but professionals usually call this realm a cognitive ecology. For a meme to survive, it needs to fit in with other memes already present in a given society; a well-crafted, hardy meme will fail to propagate if nobody in the audience understands what it refers to, just as even the most aggressive kudzu would be hard-pressed to do well in an arid desert.
Understanding a cognitive ecology means more than simply studying the target audience; it means studying the other memes the audience may be exposed to and memes that they may have rejected in the past. Memes, by their nature, alter the influence and growth of seemingly unrelated concepts. For many memeticists, the ecological metaphor is very useful, prompting them to think in terms of memetic niches, memes competing for resources, memetic parasites, even memes becoming “spores,” dropping out of sight until changes to the cultural environment trigger their return.
Cognitive ecologies are tricky. In 2103, shortly after the Pacific War, the World Trade Organization contracted Thunderbird, an Australian advertising firm, to construct a memetic campaign for Thailand. The campaign centered on encouraging the formerly nanosocialist nation to abide by the global content-rights-management (CRM) laws. Thunderhead brought in specialists in Thai culture, conducted multiple visits and interviews, and brought in test audiences for their campaign, all focusing on the myriad ways the audience would react to ads about CRM. Despite the research, it turned out that the phrasing of the message, and the characters’ appearances in the ads, echoed elements of a short-lived InVid program in the region parodying life outside the Transpacific Socialist Alliance, “General Pham Says No!” Although the show was no longer a major part of the Thai cognitive ecology, it had left enough “spores” that, once somebody made the connection, a derisive countermeme spread quickly. The ads were a dismal failure; CRM violations in Thailand actually increased over the subsequent year.
Well-crafted, complex systems of memes often form the bedrock of cognitive environments. As such, they can be remarkably hard to change. The elements in the memeplex are self-reinforcing, often structured to block countermemes and retard evolutionary variation. From the perspective of the memes’ survival, this is a good thing – persistent memeplexes can last for generations with surprisingly little change. For people who have adopted those memes, however, this is something of a problem – the challenges the persistent memeplex was originally crafted to face may no longer be relevant or even exist. Religions are the most commonly cited examples of persistent memeplexes, but ideologies and political systems often have as much or more cultural weight.
Less complex than the elaborate memeplexes at the core of an ecology, stories nonetheless play important roles. In many respects, storytelling is the oldest form of memetic engineering. For millennia, humans have created alternative worlds (or fragments of worlds) to give structure to events, emphasize social norms, and entertain. Modern meme hackers pay particular attention to storytelling, as it’s often easier to craft persuasive memes as part of a work of fiction than as a straight argument. Propaganda, conspiracies, and urban legends are often considered to be storytelling by memeticists.
Viral ideas are memes crafted to emphasize propagation, even at the expense of persistence or even stability. Viral ideas spread quickly and easily, often subverting or replacing – if only briefly – existing memes. The adoption of these memes is rarely felt deeply, however. Most viral ideas are forgotten or replaced in short order. Viral ideas are particularly powerful in chaotic memetic environments, where a rumor can rip through a mob and send it off at a new target.
Gossip is the humblest part of a society’s cognitive ecology, yet many memeticists consider it the most important element. Gossip functions as “soft” power to guide social behavior, giving everyday people a degree of power over elites. The use of gossip as part of a memetic campaign is a delicate process; the meme hacker must create a rumor that is plausible, compelling, and untraceable. If a particular line of hearsay, especially a false one, is revealed to be planted, public opinion can very quickly turn against the source of the engineered story.
It’s important to recognize that not all memes are intentional constructions. Many emerge from observations of the environment or from the intersection of multiple other memes, and propagate due to their compelling, apparently truthful, or amusing nature. In the era before memetic science, the vast majority of memes emerged spontaneously, and even those that were intentionally constructed owed their existence more to folk knowledge and ritual (memetics as alchemy) than to well-understood scientific principles (memetics as chemistry). But even in the post-PHI age, many memes appear seemingly out of nowhere, and owe their existence not to carefully designed memetic campaigns but to the human (and posthuman) ability to come up with new ideas.
People in Fourth and Fifth Wave countries at the beginning of 2155 can’t get away from seeing or hearing these ads – many can’t get them out of their heads!
1. “Mooshi Mooshi!” As said by Sarah the AstroBurger spokescow. In the ads, Sarah is dressed in a kimono while introducing the restaurant chain’s new Nattoburger – all the fun of a burger with the great taste of natto! AstroBurger claims that Sarah is a “bovine uplift,” but most people believe that she’s actually a specially designed bioroid. Children are particularly prone to screaming “mooshi mooshi” at random times.
2. “Don’t Talk With Your Mouth Full!” Accompanied by an image of a person with several dozen different food items sticking out of his mouth. Ad for SpaTek’s programmable nanopaste rations, sold in most markets under the brand name “Copia,” emphasizing the ability to select a wide assortment of food flavors. Ads usually show the unsuspecting consumer in a public setting, and in the midst of a conversation, taking a bite from a Copia bar and suddenly finding his mouth filling with different foods.
3. “Hallelujah!” The latest advertising push from Shanghai Interactive for its new MRsiv2 virtual interface glasses. Over the familiar strains of Handel’s Messiah, masses of people don the new MRsiv2 VIGs, their faces (with an appropriate diversity of ethnicity, gender, and genome) aglow with awe and happiness.
4. “The Whole World Is Watching.” A frighteningly catchy ad jingle for Durban Outfits, a South Africa-based clothing retailer. The ad is for their new line of videocloth-based clothing, including jackets, jumpsuits, and pants. Most ads finish on an image of models wearing the videocloth pants, with a second Durban Outfits ad running on their backsides.
5. “What Are Parts Without the Whole?” Ad for the upcoming InVid release Facets. They can be found across Earth and Mars, and in most medium-sized or larger space stations. The Facets campaign may be the biggest ad blitz for an InVid ever. Most show multiple images of a face crashing together and shattering.
One of the signature attributes of memes is that they aren’t static. Memes evolve over time, often over a fairly short period. In most cases, this change happens in order to make the meme fit better in various cognitive environments. In a few cases, the alteration is more deliberate.
Contrary to the sometimes gleeful, sometimes horrified predictions of turn-of-the-century pundits, increasing commerce and communication between Earth’s various regions did not lead to a single global culture; different cultural regions can and do still have strong local characteristics. Memes and memeplexes often adapt to different cognitive environments by changing important elements. For memeticists, this means that approaches and symbols that work in one part of the world may fail utterly in another. The Area Knowledge skill is useful in these situations.
Sometimes local variation provides insufficient distinction for a population unhappy with the state of the world. A growing number of people are choosing to cut themselves off entirely from others. One of the first key steps for those who live on the fringes of society (largely in space) and those who choose to be isolated (largely on Earth) is to sever many of their memetic connections to their past. This results in a fascinating combination of cultural museum and cauldron: memes that fit their situation and are under little pressure to adapt can remain remarkably stable over decades, even as their mainstream counterparts evolve to fit changing times. Memes that remain useful but are under adaptive pressure can, absent external forces moderating their changes, rapidly churn through successive variations. Outsiders visiting a well-established Isolate community often find surreal mixes of aging ideas, fashions, and jargon and utterly new, surprising descendants of established mainstream memes.
Firms that specialize in introducing cutting-edge fashions and designs to a system-wide market hungry for innovation can pay very well to get their hands on these new variant memes. Despite this, most Isolate groups avoid outside contact. Mainstream memes can be seductive, especially to those who grew up in the privation of isolation, not having lived through the identity or social crisis that triggered the move to the fringes. One of the better-known examples of this was the band “Side of Fries,” a trio of teens from an Isolate community in Alberta. Brief, surreptitious explorations of the Web gave them fragmentary samples of a variety of music, which they then combined and imitated for their own recordings – and, in turn, placed on the Free Net. Many believe that Side of Fries kicked off the “Dubcore” music trend of the late 2100s.
One of the fundamental aspects of the mainstream memetic environment in 2155 is that it is immersed in information. Nearly all of human knowledge and culture can be accessed via the Web (for a price) and most people employ wearable or implanted virtual interfaces allowing constant access to communication and information networks. This is a remarkably competitive environment for memes, as people still have limited attention that they can pay to any given idea; for a new meme to grab some of that scarce attention, it has to stand out – be more innovative, flashy, and relevant. When memes compete with each other for attention and access, memetic evolution is the inevitable result. Those memes that flourish are particularly well crafted for their environment.
A typical Fourth or Fifth Waver can access tremendous amounts of information and, with the assistance of an AI, filter it far better than people in decades past. But this constant access to information also means that blatantly deceptive or crudely manipulative memes don’t stand a chance. On Earth, a person is almost never out of reach of an answer, a critical opinion, or an alternative scenario; fact checking is remarkably simple. This doesn’t mean that deception and manipulation are impossible, only that when it is done it is usually done very well.
One effect of AI assistance is that while selection is more aggressive, random variation is far less common. Imperfect human memories are now backed up by far more accurate machines. While this means that memes retain higher fidelity during communication, it also means that happenstance improvements, where a meme is altered in the telling because it “sounds better that way,” are rare.
Virtual interfaces aren’t just used to filter out deceptive memes – they’re also used to filter out unwelcome images and opinions. Even the most basic NAI interface is able to identify and block web links and network messages that the recipient has no desire to see. Somewhat more sophisticated models can even block real-world images (and sounds, if the interfaces is implanted). Rather than seeing a lurid poster or a screamsheet commentary with retrograde politics, the person using a filter sees a blank space, or a pleasing image to let him know that offensive material has been (virtually) removed. Using filters to eliminate alternative views is generally discouraged by democratic nations, although the adoption of filters is common even in the most open societies.
Filters may occasionally be set up to alter outgoing communications, as well. A filter application called Scrubber can “clean up” text, voice, and image messages sent from a virtual interface, removing awkward or inappropriate elements, making the sender appear more educated or less offensive than he may actually be. Ideologically driven movements sometimes install Scrubber (or similar software) on member interfaces to make certain that all communications are politically pure.
One problem with traditional filters is that they are static by nature, and can be too inflexible to cope with an individual’s changing needs, environment, and cognitive ecology. Moreover, while filters may block some useless memes, they can’t find good ones. AI agents are a bit better at the latter task, and are certainly a step up from simple filtering. A good AI agent can readily track down useful information on the slimmest guidelines. But even AIs can be bogged down by the sheer abundance of material, and be unable to find good new memes as they are needed.
One solution to the abundance problem is to use “social filtering,” sometimes called “collaborative filtering.” Rather than accepting or rejecting memes wholesale, all filtering is contingent upon the source. High-reliability sources (e.g., friends or colleagues) are trusted, even when the memes they are suggesting are ones that would be otherwise unwelcome or ignored; low-reliability sources (e.g., strangers on the street or InVid “propa-tainment” shows) have to meet much higher standards when presenting new memes. In many respects, this is just how ideas have been propagated for much of human civilization – people accept new ideas from friends far more readily than they do from strangers. The difference now is the existence of virtual interfaces.
With the intimate connection of a virtual interface, social filtering can be done passively, in the background, as the interface’s AI checks with the AIs of friends and colleagues to gauge how they react to a given meme. Seconds after a person hears or reads the questionable meme, he is told by his virtual interface how the people he trusts react to it. The main drawback to social filtering is that an individual’s circle of trusted friends and colleagues is rarely large enough to be able to respond to the broad assortment of memes an individual – particularly an adventurer – may encounter. The key to keeping social filters from going stale is to entertain memes and evaluations of memes from recommended strangers.
A person’s reputation has always been important, and global information networks make it hard for an individual to escape the results of his past deeds. But for a small but influential minority, reputation is something quantified and evaluated.
In many ways reputation – specifically, the honesty and reliability of an individual – is just another meme to be passed along and evaluated. As such, memetic filters built into virtual interfaces can be used to filter on the basis of reputation. The various computer and software engineers who built the Web and wearable interfaces implemented such a system early in the 21st century for their own use. Derived from primitive reputation systems used by online games, “friend networks,” and auction sites of the time, these early reputation networks were crude, with a small number of settings that would upgrade or downgrade memetic recommendations as they were passed along. As the number of people using reputation networks grew, however, the sheer volume of trusted/mistrusted tags meant that only those with middle-of-the-road, inoffensive personalities had positive reputations; the more edgy types would typically irritate more people than they impressed, and end up with negative ratings.
The first reputation network was clunky and more distracting than compelling, and was eventually abandoned. Keeping track of affinity groups required too much human input, and the software – embedded into mobile phones, laptops, and the laughably crude wearables of the era – was too disruptive when employed. Throughout the subsequent decades, however, the idea kept popping up. Newer technologies – such as virtual interface glasses (and then implants), embedded AI assistants, highbandwidth wireless networks, and reliable face-recognition software – gave new life and feasibility to the notion of a reputation management network.
By the 2080s, the pieces were all in place. A cacophony of mutually incompatible reputation networks popped up, each experimenting with different variations of the concept. By 2098, the Working Group on Social Network Technology had drafted a standardized reputation signaling protocol, which was sent for approval to a variety of international organizations. Unfortunately, the outbreak of the Pacific War meant that few people had the time to study the proposal, and fewer companies were interested in introducing new software. Finally, in 2103, the protocol was approved and reputation management applications appeared all over the Web.
While each program has its own unique features, all reputation managers work in the same way. Rather than treat all trust ratings as more or less equivalent, the system gives much greater weight to those from an individual’s circle of friends – the more the user trusts someone, the more that user can trust the people his friend trusts. It also allows for self-defined affinity groups and filtering of ratings on the basis of trust between groups. As a result, it is possible to see not just whether a given person is trusted, but by which groups, and to see who trusts or mistrusts them. When encountering a new meme, it soon becomes second nature for those using reputation networks to quickly check the source. Who trusts this person or company, and who doesn’t?
Reputation network software is designed to be difficult to trick. Repeated entries, “logrolling” (where two or more people give each other positive entries as often as possible), mass “hate” campaigns (especially against public figures and politicians), and the like can be identified and dropped from reputation evaluations. Nonetheless, there is a cottage industry of hackers offering to boost reputation ratings artificially; few of these attempts work, at least for very long.
As of 2155, a small but growing minority of people in Europe, the PRA, and the United States rely on reputation networks to better evaluate new memes and the people or organizations passing them along. India and the South African Coalition are also starting to see reputation networks; China, the Islamic Caliphate, and the TSA have few participants. Such networks are surprisingly rare on Mars, but common on Islandia and other large orbital habitats. The smaller communities in Earth’s orbit and beyond tend not to have enough people to require AI-supported reputation management, although some are beginning to adopt reputation networks in order to have a better sense of strangers that come to visit.
Some communities may also use reputation networks as a way of managing business transactions, political affiliations, and even personal relationships. Areas that are trying the reputation society approach are generally Fifth Wave and population-dense, and always heavily networked. Notable locations with widespread participation in experimental reputation societies include San Francisco, Tokyo, Singapore, Amsterdam, and Cape Town. While not everyone participates, enough people are active in the reputation networks to make the system work. The success breeds more participation, in turn, as people who haven’t been active join in just to see how others view them. Those who participate in reputation network communities must run appropriate software. An individual may have reputation marks from dozens of different sources; there isn’t a single “reputation score.”
Most reputation networks use a generalized positive/negative axis for ratings, which range from -100 to +100. Behavior that people like typically generates positive points, often referred to as “bumps”; doing something that people don’t like results in negative points, or “dumps.” Generate enough bumps or dumps from people one interacts with, and the reputation ratings with those various groups reflect the change. An individual on a reputation network wishing to give a bump or dump to someone else on the network has to make certain that the target has been identified by his virtual interface, then give the appropriate comment, which results in a base +1 or -1 adjustment to the target’s reputation rating.
However, not everyone sees the rating the same way. How other people react to the positive and negative changes depends on how much or how little they trust the person giving the rating. A rating from someone with a very high reputation in a particular group has a far greater impact – perhaps as much as double – than one from a person with a low to middling reputation, which is typically translated into a fractional shift. A rating from someone with a negative reputation may be ignored completely, or even provoke a reverse reaction. Users are generally notified if they receive a significant positive or negative change to their reputation.
In order to prevent abuse, participants on reputation networks have a limited pool of bumps and dumps they can use. In standard networks, individuals have 10 points of rating available to use as they see fit. As those rating points are used, they recover at a pace of one point per 24-hour period. In this way, people who are careful with their rating points always have some available to respond to especially good or bad behavior. Spending this “reputation capital” on behalf of someone else is discouraged (a person really should know who he is giving a bump or dump to), and doing so for profit, if discovered, usually results in negative ratings across the board.
Newcomers to a reputation community are usually granted a zero rating; sometimes AIs scour information networks to find pertinent data about the individual in order to give a more appropriate initial score. Because of this, accusations of libel and slander are increasingly commonplace, as a negative remark on a memenet or insulting joke in a slink log can potentially affect a reputation evaluation. Although there have been libel and slander charges filed against reputation networks themselves, none have been successful; reputation systems scrupulously avoid material comments, offering only a mechanism for what amounts to entirely legal “I like him” and “I don’t like him” entries.
When viewing their own reputation ratings, most people choose to first look at the rating given by their chosen affinity group, then at the ratings given by various socially influential groups – the police, for example, or a locally dominant corporation. Most reputation management programs allow a person to see key elements leading to a particular score, including who gave various ratings; some reputation networks choose to keep that information anonymous. Even in those situations, however, the time and date of changes in reputation ratings are visible, and can be correlated to personal memory – or, more likely, an AI’s memory – of events at that point.
The vast majority of people in reputation societies have neutral or mildly positive ratings across the majority of affinity groups, although they may be highly rated within their own circle of friends. Few people are so selfless or so cruel as to generate relentlessly positive or negative reactions from those they encounter. People with higher ratings tend to be treated better, on the presumption that they will continue to behave well; those with negative ratings (especially below -50) are shunned, on the presumption that they must have done something awful to earn such a score.
By and large, reputation network ratings and a character’s Reputation map fairly well. Each positive or negative 25 points of a reputation rating is the equivalent of a +1 or -1 modifier to the reaction roll. The vast majority of affinity groups are a “small class of people” for calculating the point cost of the Reputation. (A typical person active in a reputation society will likely find his reputation rating fluctuating wildly among the various groups; at the GM’s discretion, the point costs for reputation modifiers may be ignored.) The main difference between life in a reputation network and a standard Reputation is that in areas adopting the reputation society model, the frequency of recognition is “all the time,” as other people’s virtual interfaces try to identify everyone that they encounter.