Latent Communities in an Artificial Internet

A Folklore Essay by Leo Nasskau (FC: @lsn)

There is enormous interest in building new movements today. But you don’t have to start anew if you want to change the world. Communities lie on a spectrum in terms of their origins: either they have been created to achieve a goal, or they appear to have been created almost by default; no purpose, they just exist. Who can harness the latent power of those default communities? What power do they have?


Digital Contexts

When we examined historical sources at school, my history teacher encouraged us to respect even the tiniest details. “Everything in this source was put there for a reason.”

The internet, of course, is just the same. Have you ever thought how, on the internet, nothing exists by default? Everything online was created by some person, for a reason.

Our real world is quite the opposite. Yes, the groups which make headlines are designed to effect change on the world — Google beats its earnings targets; a local charity raises money — but these Designed Communities are built on unintentional foundations.

These foundations are the product of two things: social value systems and the natural world. They arise without any purpose or design, with origins embedded so deep in the past that in the present day they appear almost as the default state of affairs.

Nobody chose them: geography that defined borders and ‘home’; birthright that influences class and ethnicity; genetics and the food chain; personality and natural ability.

It is these foundational characteristics which truly shape our lives and beliefs. They do so not just because they are part of our identity, but because they are the parts which feel hardest to give up.

This is well documented, most famously by Stuart Brand.

Some aspects of our world have higher leverage than others.
Some aspects of our world have higher leverage than others.

A community is a clearly-defined group of people whose interests are co-invested with each other. They can be built upon any of the layers that Brand uses in his diagram, whether it’s a community built around some fashion trend like Crocs or one built around a set of cultural values or beliefs like a religion.

Fashion and commerce move “fast”, he writes in The Clock of the Long Now (1999). These are the layers of those headline-grabbing Designed Communities, like Crocs, purposefully designed to pursue a specific goal.

“The fast parts get all the attention,” Brand writes. These are the parts that are easy to give up. Millennials change job every 3 years. Trends every few weeks.

The fast parts are home to a long line of changemakers who seek to influence society on a timeframe that spans only the few decades of their working lives. The art of designing a community to make a dent in the world, from startups to protests, is now highly refined.

Commercial and fashion trends appear quickly and disappear even faster.
Commercial and fashion trends appear quickly and disappear even faster.

Designed Communities have shaped our world because they are specifically designed to do so. They have institutionalised methods to accumulate assets, coordinate people, and allocate capital.

Default Communities, by contrast, have no such institutions. People forget that communities can even be formed around these ‘default’ characteristics which date to before living memory. Culture and nature are “slow”, Brand writes. They are “the work of whole peoples” in the former case and “inexorable” in the latter.

“The fast parts get all the attention. The slow parts have all the power.”

This is best seen as a feature, rather than a bug. Communities built upon these characteristics have something that Designed Communities lack: ties between their members that date to before living memory. “The slow parts have all the power.” Nation states usually beat companies.

Default Communities

Nestled in the northern English village of Trawden, the Trawden Arms made over half a million in profit last year. It helps that 120 locals run the pub for free. The pub is one of four community-owned establishments which have made the village an iconic example of what a default community can achieve.

Once a driver of the industrial revolution, Trawden’s future looked bleak by 2014. Thatcherite reforms saw the mills close down one by one, followed by the butcher’s, a baker’s, grocer’s, a post office. Come 2014 and just one of the five churches was operating, whilst the library opened just four hours a week. Buckets caught raindrops from the roof. The community centre was slated to close.

It would have done, had locals not bought the building for £1 instead. Ten years on, their cooperative manages the centre, and now the library and a shop. Seeing what the community can do, two of every three local households contributed an average of £500 each to buy the Trawden Arms in 2021 (a quarter of the region’s monthly post-tax salary).

Default—Design Spectrum

Default and Designed Communities form ends of a spectrum. Whilst members of Designed Communities are fundamentally united by self-interest, members of Default Communities are fused together by a passionate part of their identities which they share in common.

These communities lack the dynamic structure that makes Designed Communities so successful, but community leaders who can build that dynamism will find themselves leading a community that is far better placed to change the world.

Astute digital adventurers will ask ‘where are the latent communities of the internet?’ In an artificial world, every detail was created by some person, for some reason. But there are two types of characteristics we can look at, which arise essentially unplanned: unintended consequences (the ‘nature’ of the internet) and personal values (which shape its culture).

Unintended Communities

The intentional decisions which built the internet nonetheless caused unintended consequences, and it is here where our first default communities lie. Trawden’s revitalisation came as the byproduct of Britain’s most radical reforms for a generation. Default communities of the internet will likewise be grow from the most defining characteristics of our internet nature.

Specifically, there are default communities lying in wait in response to decisions made by dominant internet companies (like Amazon and Meta) and in response to the rise of new internet primitives (like search engines and AI).

Decisions made by internet companies

MENTAL HEALTH: Ignored for years by social media platforms, it is undeniable that tragic declines in mental health are driven by news feeds. Jonathan Haidt cites 476 studies to this effect in The Anxious Generation alone (whether you agree is beyond my scope here).

What this produces is a group motivated to tackle a common suffering, culminating in legislation like the UK’s Online Safety Act. In Britain, the 2017 a teenager’s suicide prompted government interest in a country already aware of social media’s manipulative potential (following the Cambridge Analytica scandal and 2016’s Brexit referendum).

Amidst the furore of British politics, which saw three elections and four Prime Ministers since that Brexit vote, the Online Safety Act was an irresistible force motivated by its own extensive volition. Buttressed by a highly motivated constituent group, both media and political voices alike then saw reason to champion the cause.

The process is well known to policy entrepreneurs. Step 1: identify an intrinsically-motivated community, or one in a crisis. Step 2: nurture its passion, shepherding it towards a specific goal. To go further, shepherd it towards a broader economy and cultural system.

Key for any default community is its source of dynamism, which makes the difference between a community like Trawden and a community which succumbs to the dynamism of others. For the mental health community, its dynamism came at first from people closely related to it: movement activists, parents of children who suffered, or social media whistleblowers. But note how other groups, from business, politics, and media, which largely picked a side later on, were critical dynamic allies in the battle against some of the most powerful companies that human society has ever produced.

Frances Haugen testifies to Congress in 2021.
Frances Haugen testifies to Congress in 2021.

INTERNET CENSORSHIP: The alliance between a community’s members and others prioritising their own interests is no more apparent than in the case of internet censorship. A movement fostered on then-Twitter, the community motivated by free speech is a blend of libertarians led by their own minds and others led by people who adopted the cause for their own gain. And the movement has had some success: anti-censorship social media platforms have reached millions of users and billions of dollars in valuations.

In large part, the growth and community of Farcaster is the product of a strong desire to develop decentralised social networks, free from the rules of today’s dominant network owners. With many thousands of daily users, Farcaster’s popularity is an example of what can happen when people are united by a common and passionate belief.

But anti-censorship has not typically been a winning cause, usually falling second-fiddle to competing interests. Whilst Elon Musk claimed he bought Twitter to make it the internet’s town square, the platform has become sharply more compliant towards government-requested censorship since his takeover.

Even the success of anti-censorship movements have mostly come from leaders who chose the cause to support other personal priorities. Donald Trump held notably mixed views on the First Amendment before becoming a free speech defender when being so was electorally advantageous. As a default community, anti-censorship was co-opted into the agendas of others, whether this benefitted the movement or not. It’s more than ironic that adverts on Truth Social are called sponsored truths.

Lessons can be learnt from both cases, but internet censorship was a particularly hard battle to win. On one hand, censorship does not rank highly on the priorities list for those who care about it. Meanwhile, those who care about it were too few in number: they went up against organisations which united billions of people. With 20 million users at Parler’s peak, even the biggest ‘free speech network’ pales in comparison.

Mapping out default communities on the internet.
Mapping out default communities on the internet.

These two limitations, the priorities of a group and its size, relatively, are key in identifying which default communities are likeliest to succeed. Relative priorities refer to how much a group’s members value the cause, compared to other priorities they have, like career progression and sharing funny TikToks. Relative size refers to how influential a group is, compared to the group from which it seeks to extract concessions (such as more stringent or more loose moderation).

Privacy campaigners also suffer from this double-limitation. Many people care about privacy to a rather low degree, and when you remove these people there are only a few who truly prioritise it. Privacy movements had victories in Europe, but these came from lawyers and policymakers, not a real community. And the opponents they faced were the same social media networks, if not even larger organisations.

Other communities are hamstrung by just one of these two limitations, with members who either are too weak in number and influence or who don’t sufficiently prioritise the cause. Groups of people who suffer from significant online accessibility issues or from internet scams are typically just too small to achieve their ambitions, despite the personal severity of their cause.

Changes in how the internet works

It is also possible for a default community to struggle even with many members, if those members don’t prioritise the cause. These are the default communities which tend to form in response to evolving internet primitives: fundamental changes in the nature of the internet.

ETERNAL SEPTEMBER: The first such change came when the online world evolved from a secluded network to the cultural zeitgeist. Writing in a time so distant that people could use their full government name online, Dave Fisher lamented on a Usenet forum that “September 1993 will go down in net.history as the September that never ended.”

He was talking about AOL’s decision to let millions of its customers access the internet for the first time ever. Internet users in the USA, numbering just 2% in 1993, more than quadrupled, ballooning to 9% by 1995.

The flood of new users overwhelmed the existing Usenet forums. And whilst disgusted oldtimers made t-shirts and text posts in protest, little else happened to protect the internet they loved. When push came to shove, Usenet was just a pastime. The internet which came next was different to Usenet, but it wasn’t bad enough to be worth seriously opposing.

INTERNET SEARCH: Almost two decades on it happened again, when the internet crossed 2 billion users in 2011. That January, TechCrunch explained Why We Desperately Need a New (and Better) Google. Of course, SEO fought battles for your eyeballs since the Usenet days, but whilst Google had originally swung the balance back towards users, SEO bait and content farms were back on top by 2011.

Two major changes soon came to the Google algorithm, called Panda and Penguin, but the appeal of monetising a customer base which has grown to many more billions means there remains a common feeling that SEO ruined the internet and Google pulled the trigger.

But who really cares?

A whole host of startups have sought to do internet search without SEO bait.

In 2011 TechCrunch sung the praises of blekko, a search engine which used human curation to find better results, and many search engines have risen up and fallen away since.

Others have also tried human curation (Sproose, Mahalo).

Thematic engines grouped results on a visual map (Kartoo, Quintura, Grokker).

Semantic tools promised to respond to your underlying intention, rather than keywords (Kosmix, hakia, Yebol).

Others tried to compete on the back-end: a bigger data set (Cuil), an open source algorithm (Wikia Search), or aggregated results from other engines.

Now is the era of the AI search: Neeva, You, Perplexity. But the lesson from two decades of failure to displace Google is that the default community which says they want an internet untouched by SEO bait… will happily accept SEO for the benefit of convenience. Google in the 2010s was different to Google in the 2000s, but it wasn’t bad enough to be worth seriously opposing.

Panda and Penguin plot how to tackle SEO bait. But AI has implications for default communities far beyond search.
Panda and Penguin plot how to tackle SEO bait. But AI has implications for default communities far beyond search.

That notwithstanding, AI is a main character all on its own. Changes in how internet content gets made strikes deeper than how it gets searched. And unlike Usenet posters or Google searchers, creators from art to entertainment have a lot to lose from computers which let anyone do their job in an instant. What comes next could be bad enough to be worth seriously opposing.

And the dynamism that default communities so often lack is present and well, in singer Holly Herndon and her partner-in-every-sense Mat Dryhurst. For years at the forefront of AI applications in music, Herndon advocates for the AI creative revolution just as passionately as she advocates for a specific path of its development: one that accommodates creators rather than running roughshod over their rights.

To identify what the pair have done well, we should remember what makes designed communities so effective: institutionalised methods to accumulate assets, coordinate people, and allocate capital. With their organisation Spawning, Herndon and Dryhurst are building these structures atop a much stronger foundation: a default community highly motivated to shape the future of computed creativity.

Spawning is most successful today as a coordination device. Their tools are the first in the world which give artists an idea about whether AI models have been unethically trained on their work, and they make it easy for artists to declare an opt-out from future model training, as well as tools to block AI scrapers.

The aim is to create an animal welfare food label for AI art models. So far, artists have protected 80 million images with the service. Whether AI companies recognise any opt-out is up to them, but Spawning’s aim is to persuade the world that stolen training data is as unethical as factory farmed meat.

Respected visual models like StabilityAI’s Stable Diffusion V3 already adhere to the opt-out, as do Hugging Face’s AI development tools, making it “effortless to honour creator wishes”. The cold start problem is out of the way.

More recently, Spawning have launched Source.Plus, a 40 million-strong dataset of opt-in and public domain imagery. Building a visual AI model that respects artists is now a choice that engineers make with a few lines of code.

The fact that Spawning’s tools represent millions of images, and are thus relevant, is testament to a massive public coordination success that likely could not have happened without the deep ties that artists have to the cause.

Having raised $3m and built a team of 6, Spawning is attempting to accumulate and allocate capital as effectively as it has coordinated people. It is easy to write a thesis explaining why it will fail, why Big Tech-backed AI juggernauts will win, and why everyone will go along with them. But the strong foundations of Spawning’s ambition make it an underrated bet that aims to change the world.

Personal Values

For Spawning and Trawden, values play second fiddle to the specific issues they work to tackle. But internet adventurers can also build Default Communities around the values themselves.

This practice is well-known as the subtlest form of advertising. Brands don’t spend millions at the Superbowl to just get 100 million eyes on their product. They advertise at the Superbowl to connect their brand with a set of values. It’s not just the fact that loads of people watch Superbowl ads. It’s the fact that everybody knows that loads of people watch Superbowl ads, and thus everybody knows the values that each brand is trying to connect with.

The attempt quickly becomes self-fulfilling as people who want to reflect that value self-select into buying the associated product. Whether it’s soft drinks or cars, the work of a brand marketer is to corner a cultural niche and then, rather than moving onto the next one (which would dilute the brand), to occupy that niche for decades.

Swoosh.
Swoosh.

The brand marketer opportunity is exponentially larger on the internet, where cultural connections spread further and faster than anywhere else. Much like the Superbowl squeezes smaller forms of marketing that can’t generate such broad cultural connections, advertising at internet gathering points will squeeze the Superbowl.

Yet whilst there are billions of people online, such internet gathering points do not really exist. Hence why Meta is investing so much in the metaverse, which is where these cultural connections will form online. And hence the excitement of Meta, Reddit, Roblox, and the like for digital assets, which are the medium for these cultural connections: over 1.5 million people adopted NFL-themed digital avatars during last year’s Superbowl, for example.

Digital assets are uniquely powerful cultural symbols for two reasons. Their owners have meaningful property rights and their provenance can be trusted. They can thus be transferred to anyone at negligible cost, and their authenticity and value can be trusted across the entire internet, not just within the walled garden of a given social network. Testifying to this, numerous organisations are working to tokenise values, culture, and authenticity, like KPMG and their partner Awsm for brand equity and Blackbird for restaurants.

However, the most common values which digital assets have been used to symbolise have so far been wealth and exclusivity. When you set an NFT as your Twitter profile picture, this is all you can achieve. By contrast: wearing your Vans at the skatepark (or on Roblox) is an authentic use of the brand, because it is a product of and a complement to your original behaviour. Compelling brands and communities cannot simply be promoted with cash: how they show up in the world matters.

Internet gathering points and digital assets allow people to authentically express their identity and values online. Gathering points are the where and digital assets are the how. As these complements come together (sooner rather than later), new values-based communities will be consecrated and grow online; digital assets will be how their members signal the values they hold dear.

Few communities have done this better than crypto projects Base and $Higher. A community built around being action-oriented, your own person, and supportive of others, Base quickly became one of the largest L2 chains on Ethereum. The clear exposition of Base’s values attracted the developers who sought to build products for the type of people those values appealed to, creating a cycle of growth which allowed Base to make a huge impact on its crypto niche, with plans to spread across the rest of the internet too.

Though a coin rather than an L2, the Higher story is much the same. “It's for all those aiming to do more, to think bigger, to leave their mark on the world,” declares the website. The community-turned-movement sponsors a group of Higher Athletes who align with the brand’s optimism and ambition. Like Vans in skate culture, wearing Higher gear is a product of and a complement to your original behaviour.

Aim Higher, on the streets of Paris.
Aim Higher, on the streets of Paris.

The Base-Higher playbook is this. Harness the power of digital assets to develop cultural connections with a set of values, and then bring dynamism to the Default Community which emerges (rather than ‘forms’) around them. This empowers community members to promote those values, where the price of membership is that this is done in a way to complement the interests of the dynamic actor (for Base: building on the Base chain; for Higher, using Higher memetic symbols).

The playbook is simpler for the Default Communities formed as unintended consequences. When a dynamic actor like Spawning forms a direct connection with the Default Community’s unifying force (like concern about AI art), the actions which that community takes are already naturally aligned to the dynamic actor’s interests.

But the approach of recognising these deeper sources of connection, which people already share, is common to both playbooks. Today, there is enormous interest in building new communities, but it is not the case that if you want to change the world, you have to start anew. Build where passion already exists, and is latent, waiting to be awakened.


Further reading:

  • The Pace Layers of Civilization, notes on Stewart Brand’s book by Caterina Fake.

  • Stuart Brand interviewed by Palladium Mag in 2022.

  • Ads Don’t Work That Way, by Kevin Simler.

  • Aim Higher on the streets of Paris, photo via @kugusha.

  • Moral Ecosystems, by Toby Shorin.


Footnote – in the middle of the Default-Design Spectrum

In the middle of the Default—Design spectrum lie communities which combine elements of both sides. Typically, members of these middle communities form ties around shared passions but with no shared purpose. But because coincidental communities tend to form around specific ideas, they can adopt a shared purpose when motivated by leaders linked to those ideas.

Their membership is a disinterested impulse, to borrow de Tocqueville’s term. In using this term, de Tocqueville does not mean that the participants are not interested in the community of which they are a member. My interpretation is that he means they participate in the community from their personal passion (ie, an impulse), and do so without thinking about the broader interests of that passion or community (eg, how they might support it politically or elsewhere in the public arena).

A modern example is politically-active K-pop fans, with community bonds forged by a love of music that have been deployed to achieve very different goals. Coincidental K-pop communities are why journalists have found themselves exploring questions like How K-pop fans became a political force to be reckoned with. That headline came in 2020, when fans of BTS registered tickets at a Donald Trump rally en masse, only to not show up and leave swathes of empty seats.

Subscribe to Folklore
Receive the latest updates directly to your inbox.
Mint this entry as an NFT to add it to your collection.
Verification
This entry has been permanently stored onchain and signed by its creator.