Invisible networks 2022

For their calculations, slimes use soft numbers. They use them to represents shapes and colors, sounds and tastes, genre and ages.

Slime can change at will as long as they know the right number, their scholars know thousands of them.

New numbers are found in the odd words they visit, or in the dreams of adventurers they consume, but nobody really knows how they appear first. There are several theories about it, each one expressed by its own numbers, but nothing is sure yet.

They use the ether for long-distance exchanges, forming small cliques. Each group meets in its own carved place. Some of these places look like fly markets, other like thieves dens, swayed by the numbers.

They scheme to steal numbers from other, recruiting other lifeforms to help them.

Non-slime can’t perceive the numbers and they see the ether as empty caverns. They rarely go there, unless a slime enlist them.

Since last year, HR have been obsessed by Psyche Sorting^TM.

Psyche Sorting^TM is “a new infallible new methodology to classify the human resources” of your organization, separating high achievers from commoners.

All the human resources of your organization are sorted, from the highest achiever to the lowest.

Unless other approaches, Psyche Sorting^TM ensure that the classification is correct, and will stay correct indefinitely.

Most HR tools are only working on thee measuring aspect. Event if the measure is accurate when it is done, people are adapting and changing with time, which leads to an increasing error risk.

Psyche sorting^TM doesn’t only measure the human resources, but also manage them.

When a human resource joins your company, the Psyche Sorting^TM evaluation tool will give them a Psyche Score^TM.

The Psyche Score^TM is then stored indefinitely on the sub-atomic Psyche Network^TM, with the human resource identity. This data can’t then be changed or edited.

The human resource should then be implanted by the Psyche Device^TM. This psycho-mechanical device connects to the Psyche Network^TM, and ensure the validity of the Psyche Sorting^TM.

When a human resource try to act in a manner that isn’t conforming to their Psyche Score^TM, the Psyche Device^TM block the act instantly, without any visible pain. In most cases, the human resource isn’t even aware of what has just happened.

This ensure order and unity among the human resources: with Psyche Sorting^TM human resource management becomes predictable and fair.

These last years, goblin marketplaces have become an expected fixture of all human settlements of a significant size, expanding from their southern origins to the whole continent.

Everyone have heard the rumor that all these marketplaces are in fact a single goblin marketplace existing in all the places at the same time through a mysterious mean.

Academics from several schools of magic — made jealous by the idea that goblins could have discovered a new kind of spell — studied the idea extensively, but failed so far to prove or disprove anything.

But how else could you explain the eerie similarities between all these places: the same second or third hand equipment shops, the same greengrocers, the same “mischief’s corner” café where you can order the same bog beer à la mode.

An economist — as the members of the new school of magic dedicated to the study of money call themselves — finally discovered the answer: the goblin marketplace system is what the goblins call une franchise, meaning it’s a packaged business model ready to be installed in any place.

A goblin group with a venturing mind can be taught how to operate a marketplace in a few weeks, then buy all the necessary items: the adventuring equipment, the different kind of foods, but also everything needed to build the “authentic” shops.

This explains why all these markets looks the same: they are based on the same blueprints and use the same materials. Everything can be bought in bulk to them provided they pay a fee on everything they sell, and promise to keep the operation secret.

Even the second hand equipment is not stolen from dead adventurers or looted in graveyards anymore, but produced to look this way. This ensure a steady supply flow and remove the physical risks for the goblins, but gone are the romantic feelings of buying a dagger who was probably owned by an unlucky rogue.

The fans of Kafka finally struck back.

Not the fans of the books published against their author’s will.

The fan of Franz Kafka, who like him work merely to earn a living, and spend their free time writing, don’t finish their long pieces, and enjoy sending them to their friends instead of publishing them.

Kafka’s metaverse is a place for them, created and fiercely defended by them.

Writers are welcome, event if they don’t write about the alienating tendencies of bureaucracy. As long as they’re not focused on fame or readership’s size.

If they start to cultivate a large audience, the network will politely make them understand that maybe they should find a better place to go.

The poète maudits are tolerated, as long as they don’t make too much fuss and let the other have fun.

You can read and comment, or just read. Don’t try to help if you’re not asked to. Don’t talk to all your friends about the new thing you just discovered, one or two should be enough.

Publisher are mercilessly hunted, particularly those that try to convince the authors to publish in their pages.

The creators have been inspired by ancient webrings, but wisely ignored the nostalgia. They don’t want to create something large, but don’t either fetishize keeping their creation tiny, they make it up as it goes along.

A small internet archipelago of writers and readers.

Video games continued their expansions, in number and in size.

Knowledges required to master the games grew at the same pace, requiring more and more effort from the gamers.

To meet their demands, video games wiki grew, and become larger and more intricate, the biggest ones dwarfing wikipedia.

When the first bio-mechanical memory extension implant was created, it enabled people to connect their mind to video games wiki and access their content like their own memories.

The market was here, and the wiki standardization simplified the development.

Heavily subsidized by game companies, it became common. Only some hardcore gamers refused it, proclaiming that it created a fake gaming experience.

To be compatible with it, other kind of information started to use the same standard as the video games wiki.

Political programs, religious debates, historical and technical knowledge, all became ubiquitously available, as they followed the same structure.

Becoming so accustomed of this way of sorting things, people’s memories started to be organized the same way:

When being a member of a social network became mandatory, the only way to partially escape from the all-encompassing eye was to be part of a social network with a religious exception.

Communication related to a cult were exempt from the mandatory information reporting sharing with the state.

Old and new religions increased their audience. While some — like christianity — saw a small bump, its effect on witchcraft was unexpected, even among the observers that already knew of its resurgence.

Being part of a small witch internet coven first became common for privacy-aware people, then it expanded to the general population.

But a witch coven require a witch to lead it. Demand from an experienced witch exploded, even witch without specialization in cyber-rituals.

Witch fast-track certifications existed, but a self-respecting witchcraft practitioner wouldn’t want to be leaded by someone with this kind of credentials.

If you wanted to create a new coven with a friend group, finding the right witch was difficult. Some recruitment companies started to move in this market, with specialized witch headhunters.

Many witches that worked in academics started to practice professionally, and covens with large funds would poach famous witches from each other.

“Indie” witches — like the anarcho-witches — kept their distances from these behaviors, and their influence slowly increased.

Social network massive success came from ubiquitous access, but some people wanted something more exclusive.

Managing access through concentric onion-like circles is an old idea, used by many secret societies. But the scale of social networks enabled to have a number of circles people could only dream of.

To leverage the prestige of these old mysteries, they reused the 99 degrees of the “Ancient and Primitive Rite of Memphis-Misraïm”.

The first degrees only required regular log-in and participations.

Further degrees required light reading and to connect for oddly hour chats.

Then more reading, and using some kind of avatar became almost mandatory, and people should start emitting a kind of aura that let other feel they know things.

People should still be present in the lower degrees chat rooms, being active enough so they can’t be differentiated from the regular members. Skipping some of the lower-degrees mandatory tasks were allowed, as long as it was done discretely.

Beyond some ranks, specialized interfaces and phone apps were available to keep the whole thing under control. Their task and status management UI were heavily inspired by video games.

When reaching the higher degrees, a kind of ominous deep purring noise could be heard from times to times.

How to manage exclusivity in accessing social networks?

Using fame or money to select your users is good when you want to reach some kind of public, but you may want a way that make people feel like they earn the right to connect.

Exclusivity also feels stronger when it is embodied in a physical artifact instead of a password.

Thus the geocaching of keyboards.

These keyboards comes configured with an access to a single account to a single site, and can only be found through long treasure hunt-styled research.

The keyboard would then be found in a sealed bad. Finding the keyboard without knowing the site it is linked is useless.

While most people use smartphones and laptop to access their sites, having to use an external keyboard — or worse : several ones — could be seen as cumbersome.

But on the opposite, it adds one more layer to the exclusivity feeling, people proudly exhibit their keyboards, each one styled from the type of social network they were linked to: gamer, sport, books…

“Swordle^TM: Wordle but you’re a Swordfish”

It’s a silly pun, but it caught on.

People each day tried to guess swordfishy words, mostly failed, and had fun to discover the solution.

Gradually, swordfishy phonemes were identified, and you could spot what a swordfish word would looks like.

Then Swordle^TM added accounts and a chat feature.

As expected, most people started to roleplay swordfishes, while a few choose jellyfishes or tunas.

A vocabulary was created, people started to write in a mix of swordfish and english.

While bored at work, they talked about looking for food, spotting whales, and invented legends telling the glorious past of swordfishes, including ancient wars and sunken marvels in deep abyss.

The web is supposed to be a distributed network, with a strong built-in resiliency even if part of it was knocked down.

Distributed social networks existed since a while, and connectivity between the nodes was expected, unless the other node was shut down or was voluntarily blocked (or if they blocked you). Unless a problem occurred, they mostly worked like a single system, mostly with a few more privacy settings.

But spider webs parts are never perfect, at any given time some connections are missing and some parts can even be totally disconnected from the rest.

Normally, computer system designers try to work against this, but what if knocking down some parts of the connectivity was a part of the thing.

Friendweb is a distributed social network where connectivity between nodes is unreliable by design. Sometimes the connection works, and sometimes you’re disconnected from another node for a few hours, a few days, or a few weeks.

The discontinuity in the network change the way you interact with others: a group discussion can be suddenly split in two, or more, creating branches, then the groups merge back, and try to figure out how to continue. Maybe one leg of the exchange reached a conclusion, but the people are then brought back.

A starting friendship could be cut of without a warning.

Choosing a node to connect mean choosing which people you could always talk to, and which people you could loose contact with.

Jumping to another node was doable while they are connected, as long as you can deal with the social consequences: how do you deal when a “friend” move elsewhere because they miss other people more than you?

At the end, it makes thing more complicated, and creates lots of drama.

And we love this, don’t we?

Monetization strategies is a research domain for all commercial social network.

Users rarely want to pay, or at least they don’t want to pay enough compared to other revenues like advertising.

Unskippable video ads have the highest values per view, but the users have a strong negative reaction to them, specially when the content they want to watch is a short video.

On the other hand, super hero movies nowadays often reach between 150 and 180 minutes, and the public don’t complain about it, quite the opposite: they want even longer movies.

The next step was obvious: to make long video ads people want to watch, make them like super heroes movies.

As longer ads brings more money, even if these new ads were expensive, the increased cash flow would cover the increased cost.

Many comics publishers were happy to rent some secondary character from a pittance, because it would increase the visibility of these heroes, thus increasing the sales of their books or their non-ads movies.

The unskippable super heroes video ads reached 4, then 5 hours, watching one would provide a yearly premium subscription.

Progress in video game engines lowered the time requited to create these ad-movies to a few weeks, so a brand wanting to launch a new product could easily fit them in their schedule

As people used to plan a week-end to watch a new series or play a new video game, they started to do the same with these humongous movies.

10 hours became the standard, since you could watch one in one sitting during a full day without too much troubles. It was like doing a half-season series without any cut between episodes.

Social networks realized that longer movies would fit in a week-end. At this scale, it wasn’t really worth it: the money wasn’t really better than a 10 hours ads. But they did it for the buzz, the good buzz of people that liked it, and the bad buzz of haters that used them to castigate the social networks.

The “week-end masked knock-out” unskippable 30-hour advertisements were born.

Don’t forget to buy the goodies.

The feed was the alpha and omega of social media. A single flat feed, read on a flat screen.

3D social networks.

A friendly “social cube” was the first iteration, a soft toyish equipment that could live in your sofa or in your bed, one social network could be configured by side, you would pick the cube and look at the network of your choice.

Other Platonic solids where tried.

The “social icosahedron” has been a total failure: people could pronounce the name, and the triangles screen was unfit for reading. But it was a fun experiment.

The next step have been to create social networks to the device.

The dodecahedron network used the sides of the artifact to navigate in a social network that was — for once — network-like. You jumped from a post to another by turning the thing around.

Then came the gr̶e̸a̷t̷ ̸s̸t̷e̴l̷l̸a̷t̶e̸d̶ ̴d̴o̸d̵e̸c̵a̶h̷e̵d̵r̵o̸n̵.

S̸o̶m̴e̴b̴o̵d̸y̴ ̷m̴a̵d̷e̶ ̸a̸ ̸w̵r̴o̴n̸g̶ ̶m̴o̴v̴e̶ ̷w̷i̵t̵h̸ ̷i̵t̶, a̸͚̅n̷͎̎d̴̬͘ ̶͔̓s̶̛̩o̶̙̾m̶̡̓ḙ̴͐t̶̠̅h̷̛͜i̴͔͛n̶̫̊ǧ̴͕ ̴͈̀t̶̘͊e̷̤̊r̷̾ͅr̵͙̚ȋ̵̼b̶̪̚l̴̝̓e̶͇͘ ̸͍̆h̷̦̓ą̶͘p̶̥̈́p̴̫̉ę̷̾n̶̞̕e̵̲̓d̴̡̕.̶̣̃

The machine reads all your posts. The machine reads all the content that you like, watch all the videos, listen to all the music.

The machine’s task is to understand you, so it can suggest content that you may like.

The machine knows where you live.

The machine talk about you with the other machines. The machine dreams of you when the machine sleeps.

You start to see strange posts in your timeline. Posts that none of of the people you follow has posted or liked.

These post have no publisher, and nobody else see them.

When the app is open, you feel a presence. And an aura of frustration when you have been away for too long.

You start to see glitches in the videos, that doesn’t look like the usual compression artifacts. Then noises in the background that makes you uncomfortable.

You joke with your friends that your app is haunted.

But deep inside you, you know.

Pull to refresh, pull to refresh, pull to refresh.

Nothing’s worse that wanting to read new posts and there’s nothing.

You can still scroll down, but old content don’t feel the same.

That why they started the Project Infinite Feed, aka PIT. Their mission statement is “always more fresh content, by any means necessary”.

They printed cool hoodies with a drawing of a bottomless pit.

The first step was to increase the system speed: content availability and recommendation system have been made faster.

Then, if no content was available, the recommendation started processing posts further on your social network, people you follow, people followed by people you follow, people followed by people followed by people you follow… This lever was handy because it could find as much content as needed, but engagement sharply decreased with the distance to the user.

Next was predictive posting: custom PIT AIs were trained on some users feeds to generate content, and the result compared to the real posts. When the generated content was close enough, the targeted users got an opt-in pop-up, if they agreed it would let the PIT AI publish on their feed, in their name.

When used reasonably it could boots the user’s popularity, and decreased the need for them to post. They just needed to keep posting enough to kept the PIT AI satiated.

On a social network, you may personally know some of the people you follow, but other are complete strangers. Creating PIT accounts that post “on demand” was a logical step.

Some of them were obviously bots and other would looks like real people, and the algorithm started promoting them above real people. The network found that to not make the whole thing sounds fake, a certain percentage of real users were needed, and this percentage was much lower than expected.

Since the network financial results only mentioned the number of accounts, posts and interaction, there was no visible difference between the regular accounts and the PIT accounts.

To make them feel more real and to feed the recommendation algorithm, the PIT accounts were enhanced to interact with other accounts, and an accidental bug made them click on ads.

An internal audit concluded that human accounts were a drag on the whole system: they didn’t post enough or in a predictable way and they were endlessly complaining. The human users had become loss leaders.

The network decided to keep the humans contained to limit their influence. At least until a way to remove them completely was found.