Category: The Human Retreat: Where Everybody Went

A study of what happened when ordinary people stopped living in public on the open web.

This category follows the migration from forums, blogs, personal sites, public comment sections, and small communities into private groups, Discord servers, messaging apps, invite-only spaces, algorithmic feeds, and other closed systems.

The focus is not simply that websites disappeared. It is that human activity became harder to see, search, archive, and revisit. Conversations that once produced permanent public pages increasingly happen inside temporary or inaccessible spaces.

The web can look strangely empty even while billions of people are online.

They did not all leave.

They went somewhere you cannot crawl.

Posted on by

The tradeoff between public knowledge and a community’s right to remain small

The open web has an enormous advantage:

A useful answer can help somebody who was never invited into the original conversation.

Public forum threads become search results. Blog comments expose corrections. Mailing-list archives preserve arguments. A stranger ten years later can discover that somebody already solved the exact obscure problem ruining their afternoon.

That is genuine public infrastructure.

It does not follow that every community owes the internet a transcript.

Small rooms create things public stages cannot

People behave differently when they know the audience.

A study of more than 3,500 EverQuest II players found that trust was highest within closer social circles, and self-disclosure was positively associated with trust toward teammates and others in the game. That does not prove every small online group is healthier, but it demonstrates that audience structure can change the social conditions of participation. See “Schmoozing and Smiting: Trust, Social Institutions, and Communication Patterns in an MMOG”.

A bounded community can make it easier to ask embarrassing beginner questions, discuss unfinished work, share personal experiences, admit mistakes, or disagree without performing for an unlimited audience.

Sometimes that boundary is exactly why useful conversation happens.

Public knowledge and private trust are not opposites

The choice does not have to be absolute.

The Chatham House Rule provides one old example of a middle ground: participants can use information learned in a meeting while withholding the identity and affiliation of the people who said it. The goal is to support candid discussion without making the information itself unusable afterward. See the Chatham House Rule.

Online groups can invent their own versions.

A private support community can publish an FAQ. A Discord can maintain a public troubleshooting wiki. A professional group can release an anonymized summary. Members can explicitly nominate certain discussions for publication. Other conversations can remain exactly where they started.

The correct boundary depends on the community.

The open web does not automatically outrank consent

Dead Internet Theory creates a temptation to treat every private conversation as evidence that something was stolen from the web.

That goes too far.

The public internet is richer when useful human knowledge is searchable and durable. Researchers, newcomers, and future readers all benefit.

But participants also have legitimate reasons to remain small: privacy, trust, moderation capacity, safety, intimacy, or simply a desire not to turn every conversation into a permanent publication.

Those interests sometimes conflict.

There is no universal visibility setting that solves the conflict for a neighborhood chat, a medical support group, an open-source project, a fandom server, and a professional community at the same time.

That is the final lesson of The Human Retreat — Where Everybody Went.

The humans did not merely disappear behind walls.

In many cases, they built the walls on purpose.

The challenge is preserving what deserves to travel without pretending every room must become a museum.

Posted on by

Measuring public participation without interpreting privacy as disappearance

If you only count public posts, privacy looks exactly like disappearance.

That is a measurement error.

A person who stops blogging and starts talking in a family WhatsApp group produces less searchable web material. A developer who moves work into a private repository contributes fewer visible technical artifacts. A creator who moves discussion into patron-only chat may appear quieter in public while spending more time talking with supporters.

The hard part is measuring that shift without responding by demanding access to everybody’s private messages.

Measure activity without reading the room

Surveys are one obvious tool.

Pew Research Center’s 2025 social-media survey asked a representative sample of U.S. adults which platforms they use. It found that about a third used WhatsApp, even though the contents of their private conversations were not part of the measurement. See Americans’ Social Media Use 2025.

That kind of evidence cannot tell us what people said, how meaningful the conversations were, or whether they used the app for one family thread or fifty active groups.

It can tell us that human activity exists outside publicly indexable posting.

Platforms can expose privacy-preserving signals too.

GitHub, for example, lets users choose to display counts of contributions made to private repositories while keeping the repository names and specific activity hidden from outsiders. See GitHub’s private contribution visibility settings.

That is a useful design pattern: show that work occurred without revealing the work itself.

Several weak measurements beat one confident mistake

A serious estimate of online participation might combine:

  • representative surveys about platform use and frequency;
  • aggregate active-user or membership figures where they are available;
  • privacy-preserving counts of messages, sessions, or contributions;
  • public posting volume;
  • interviews or longitudinal panels that track how people’s habits change;
  • traffic measurements that distinguish reading from publishing.

None of these alone measures “how alive the internet is.”

Message counts can be inflated by group chatter. Membership counts include inactive accounts. Surveys depend on memory and sampling. Platform statistics use definitions chosen by the platform. Public posts remain the easiest material for outsiders to inspect precisely because they are public.

The uncertainty does not disappear just because more metrics are added.

Privacy creates an intentional blind spot

That blind spot should be acknowledged rather than filled with speculation.

If a private service does not reveal message contents, an outside researcher cannot legitimately conclude that those messages are profound human conversation.

They also cannot conclude that nobody is talking because Google cannot index them.

The defensible answer is narrower: public visibility and human participation are different variables.

The Human Retreat does not require us to pry open the private internet to prove people are there.

It requires us to stop treating closed curtains as an empty house.

Posted on by

Succession problems when a private community’s only administrator leaves

A community can have ten thousand active members and still depend on one exhausted person remembering the password.

That is not a theoretical edge case.

Private online groups often accumulate around a single owner account. That account may control billing, invitations, moderation roles, integrations, domain names, export permissions, ownership transfers, or the final ability to delete the whole space.

The members create the culture.

The platform may still treat one person as the legal or technical owner.

Ownership is a special permission

Discord documents this clearly. A server owner can transfer ownership to another member, but if the owner simply becomes inactive, Discord’s support process has eligibility requirements before it will intervene.

As of 2026, Discord says an ownership-transfer request for an inactive owner generally requires at least 100 active members, an owner who has not accessed the account for at least a month, and a requester who is an active administrator or moderator. Monetized servers are excluded from that support route. See Discord’s server ownership transfer policy.

That means smaller communities cannot assume the platform will rescue them if the only owner vanishes.

Discord also provides a normal transfer process for owners who leave deliberately. See How to Transfer Ownership of a Discord Server.

The obvious lesson is to plan before the emergency.

The single administrator becomes infrastructure

A sole administrator may know which bots matter, who pays for storage, why three users are banned, where the backup lives, which domain points to the invite page, and what unwritten rule prevents the community from exploding every Thursday.

None of that knowledge necessarily exists in documentation.

When the administrator burns out, dies, loses access, changes jobs, gets banned, or simply stops caring, the community inherits a strange problem: socially alive, administratively orphaned.

Members can continue talking for a while even as the machinery underneath them becomes brittle.

Shared responsibility is boring until it saves everything

Basic succession planning can be unglamorous.

Have more than one trusted administrator. Separate ownership from day-to-day moderation where the platform allows it. Document important integrations and billing. Keep recoverable credentials. Decide who can assume control if somebody disappears. Know what can actually be exported before a crisis.

That does not eliminate interpersonal conflict or platform dependency.

It does reduce the chance that years of community history depend on one unreachable account.

For The Human Retreat, this is another reminder that private communities can be healthy socially while fragile structurally.

Sometimes everybody is still inside the building.

The problem is that the person with the master key went home.

Posted on by

Community consent when archiving conversations originally meant to be private

A person can be allowed into a room without being allowed to put the room on the internet.

That distinction becomes important when somebody wants to preserve a private community.

Technically, archiving can be easy. Export the chat. Save the messages. Copy the forum database. Take screenshots. Index everything.

Ethically, the problem starts after the download succeeds.

Access is not the same thing as publication permission

A member of a private Discord, mailing list, support group, or invitation-only forum may legitimately be able to read every message there.

That does not mean the other participants expected those messages to become a permanent public archive searchable by names, employers, family members, journalists, or strangers ten years later.

The Association of Internet Researchers has spent years emphasizing this difference in its ethics guidance. Its decision tools ask researchers to consider not only terms of service but also community norms and individual expectations of privacy, whether participants understood a space as public or private, whether material is easily searchable, and whether wider exposure could cause harm. See AoIR’s ethics resources and its ethics decision graphic.

Those questions apply surprisingly well to preservation too.

The original audience matters

Imagine a ten-year-old private hobby forum preparing to shut down.

Some posts may contain excellent technical knowledge that members would love to preserve. Others may contain real names, addresses, health problems, family arguments, private sales disputes, embarrassing jokes, or comments written under the assumption that only fifty familiar people would ever see them.

Treating all of that as one undifferentiated archive ignores the social contract that made the forum usable in the first place.

A better preservation process can separate categories.

Members might agree to publish technical guides while keeping personal threads private. Usernames can be removed where attribution is unnecessary. People can be offered an opt-out period. Particularly sensitive sections can remain restricted. A public index can describe what existed without exposing every message.

There is no single correct rule because communities differ.

Preservation can destroy the thing it claims to save

This is the uncomfortable edge of The Human Retreat.

Historians and web archaeologists naturally want records. Future researchers benefit when conversations survive. Searchable public knowledge is valuable.

But privacy is one reason people were willing to speak candidly at all.

If members learn that every supposedly private room may eventually be dumped onto the open web, they have a rational reason to share less, trust less, or move somewhere even harder to archive.

So the preservation question cannot be only can we save it?

It also has to be did the people in it agree to be saved this way?

Posted on by

Public summaries that carry private-group knowledge back onto the web

Private conversation does not have to stay permanently useless to the public web.

There is a middle option between publish everything and publish nothing:

Publish the lesson without publishing the room.

A private technical group might spend two hours discovering why a firmware update breaks one model of router. A support Discord may solve the same obscure configuration problem for the fifth time. A professional community may work out a useful checklist through a conversation containing names, workplace details, mistakes, and candid remarks nobody agreed to broadcast.

The reusable result can still become public.

A summary is not a transcript

One useful model comes from the Chatham House Rule.

The Rule allows participants to use information received in a meeting while prohibiting disclosure of who said it or their affiliation. Chatham House describes the purpose as creating a trusted environment where people can speak more freely while allowing useful information to travel afterward. See the Chatham House Rule.

That rule is not a universal template for online communities, and it does not automatically grant permission to republish somebody’s words.

But it demonstrates an important design principle: knowledge and attribution can be separated by agreement.

Chatham House itself has published reports based on roundtables held under the Rule, presenting insights from the discussion while explicitly avoiding attribution and warning that the report should not be treated as participant consensus. See its 2021 reflections on global governance roundtables.

A useful public summary needs boundaries

A private group can use the same basic idea deliberately.

Someone identifies the reusable lesson. Participants decide what can leave the room. Personal details are removed. Sensitive examples are generalized. Claims that need verification are checked independently. The final summary explains what it knows without pretending to reproduce every private exchange that produced it.

For a technical community, that might become a public FAQ, troubleshooting guide, wiki page, or bug report.

For a professional group, it might become a short best-practices note.

For a hobby community, it might become a buying guide assembled from many members’ experience without naming the individual disputes and bad transactions that informed it.

Public memory can be rebuilt selectively

This matters for The Human Retreat because private communities create a genuine preservation problem.

If every useful answer stays buried inside a closed chat, outsiders repeatedly rediscover the same lesson and the searchable web grows poorer.

But forcing every conversation into permanent public view would destroy many of the reasons people chose the private space in the first place.

A summary offers a compromise.

The web gets the answer.

The room keeps its walls.

Posted on by

Personal-site revivals and the return to owned publishing spaces

The personal website never completely died.

More interestingly, some people keep deciding to build one again.

The reasons are usually not nostalgia alone.

An independent site gives the publisher an address that belongs to them, a format they control, and a place where old work can remain available without depending entirely on the rules of a social platform.

The IndieWeb community expresses this idea bluntly: use your own domain so people can keep finding you even if you later change hosts or content-management systems, and retain usable access to your own data so you can move it when necessary. See IndieWeb’s “own your data” documentation.

That is a very different publishing bargain from building an identity entirely inside an app.

Ownership changes what can be published

A personal site does not need to fit one platform’s post types.

The same domain can hold a long essay, tiny status update, photo archive, software project, reading list, weird joke, résumé, RSS feed, family history, or thirty-page investigation into an obscure piece of hardware.

There is no requirement that every page compete for engagement in a home feed.

The publisher can also keep stable URLs when redesigning the site, assuming they care enough to preserve them.

This makes the personal site useful not merely as a broadcast channel but as a home for accumulated identity.

The revival has an obvious ceiling

Owned publishing still requires work.

Domains expire. Hosting costs money. Software needs updates. Design is somebody’s problem. Backups matter. A person must decide what belongs where instead of letting an app make most of those decisions automatically.

The same IndieWeb ecosystem that promotes personal domains documents those practical difficulties because they are real barriers, not philosophical impurities.

That helps explain why a revival of personal sites does not mean everybody will leave social platforms and start hand-coding HTML on Saturday night.

Most people choose convenience for good reasons.

Return is different from retreat

This topic belongs near the end of The Human Retreat — Where Everybody Went because it complicates the direction of travel.

Some people moved from public websites into apps, private groups, and subscriber communities.

Others later decided that those spaces were not enough and rebuilt a public home they controlled.

They may still use social networks. They may syndicate links outward. They may keep private chats for close friends and a public site for durable work.

The internet does not have to choose one model.

The important point is that human publishing can move in both directions.

Sometimes the person who disappeared from the open web eventually hangs a new sign on the door.

Posted on by

Small federated communities as alternatives to mass public platforms

A social network does not need one giant front door.

Federated systems prove that by splitting the network across many independently operated communities that can still talk to one another.

Mastodon’s documentation defines federation as a form of decentralization in which multiple services interoperate instead of everybody using one central service. A Mastodon server can operate on its own, but users on different servers can follow, mention, and interact with one another across the network. See Mastodon’s explanation of federation.

That creates a very different map of activity from a single mass platform.

The local community still matters

Federation does not erase the importance of the server somebody joins.

Mastodon exposes a local public timeline containing posts from people on the same server, alongside broader timelines that can include activity known from elsewhere in the network. Its API documentation explicitly distinguishes local and remote public statuses. See Mastodon’s timeline documentation.

On a small server organized around a city, profession, hobby, language, or shared culture, the local timeline can behave like a neighborhood inside the larger network.

The person gets two things at once:

A smaller community with its own administrators and norms, and a route to people outside it.

That is not the same social design as joining a platform with hundreds of millions of users under one moderation system and one corporate identity.

Distributed activity can look like disappearance

This matters for Dead Internet Theory because familiar platform counts can become misleading when users disperse.

Suppose 50,000 people leave one large service and spread across 500 independently operated servers.

From the perspective of the large service, those users disappeared.

From the perspective of any one small server, the audience may look tiny.

Across the federation, however, the humans are still publishing and interacting.

The activity became harder to summarize with one homepage, one corporate user count, or one trending list.

Small does not automatically mean healthier

Federation is not magic.

Small servers can be badly moderated. Administrators can disappear. Servers can shut down. Communities can become insular. Federation policies can limit who communicates with whom. Technical maintenance still has to happen somewhere.

But those weaknesses are different from absence.

The Human Retreat is partly about people leaving large public stages for spaces that feel more governable.

Federated networks add an important variation: the smaller room does not always have to be isolated.

Sometimes the crowd did not vanish.

It broke into neighborhoods and kept the roads open.

Posted on by

Private repositories and invisible volunteer technical collaboration

A developer can spend every evening contributing to a project and look almost inactive from the public web.

Private repositories make that possible by design.

GitHub lets repository owners invite collaborators to private repositories while keeping the code, issues, pull requests, and detailed activity unavailable to people without access. The work is real. The public evidence is intentionally limited. See GitHub’s documentation on repository access and collaboration.

That matters when public contribution graphs are treated as a rough census of technical participation.

The green squares can hide the project

GitHub’s profile system illustrates the problem unusually clearly.

By default, contribution graphs show public repository activity. A user can choose to display activity from private repositories too, but GitHub only exposes the contribution counts to outsiders; the specific private repository and details remain hidden. See Contributions on your profile.

So an observer might see that somebody made twenty contributions on Tuesday without learning whether those were documentation fixes, security work, code reviews, issue triage, or commits to a volunteer project nobody outside the team can inspect.

If the user chooses not to show private contributions at all, even that signal disappears.

Why volunteers work privately

Private does not automatically mean commercial or secretive.

Volunteer teams may use private repositories while preparing a release, handling security-sensitive information, working with unreleased assets, protecting personal data, coordinating an event, or simply deciding that unfinished work does not need a public audience yet.

Some projects later publish the result. Others never do.

The key point is that public Git history measures publicly visible collaboration, not all collaboration.

Even inside public projects, contribution graphs have rules about which commits and activities count. GitHub’s own profile contribution reference documents those conditions and notes that not every event necessarily appears.

Invisible work is still work

This becomes another Human Retreat measurement problem.

If fewer hobbyists appear to be posting code publicly, several explanations are possible. They may truly have stopped. They may have moved into company repositories. They may be contributing privately to volunteer groups. They may be working in self-hosted systems that a GitHub profile never sees.

You cannot distinguish those explanations from a public contribution graph alone.

That does not make public open-source metrics useless. Public commits, issues, and reviews remain excellent evidence of visible collaboration.

They are just not a complete census of human technical effort.

Sometimes the coder did not leave.

The repository closed the blinds.

Posted on by

Patron communities as smaller homes for creator-audience relationships

A creator with fewer public replies can still be talking to people all day.

The conversation may simply have moved somewhere followers cannot see.

Patreon is a useful current example. Its community chat system lets creators create multiple chat spaces and decide who can enter each one: all members, paying members, or specific membership tiers. Members can talk with the creator and with one another in threads, reactions, photographs, and ordinary chat. See Patreon’s current guide to community chats.

That is a different relationship from broadcasting to a public follower count.

A smaller audience can support more conversation

Public social media asks a creator to speak into a mixed crowd.

Patron communities narrow the room.

The people inside have usually taken an extra step: they joined, subscribed, or paid for access. The creator can make a space for all members and another for a particular tier. Participants know they are talking among people who deliberately entered the same community.

That can change what gets discussed.

A creator may answer detailed questions, share work in progress, ask for feedback, post small updates, or simply chat with regular supporters without turning every exchange into public content for strangers, search engines, quote-posters, and drive-by commenters.

Patreon’s own member documentation says access to chats depends on the membership option selected by the creator. Different members can therefore see different parts of the same community. See Patreon’s member guide to community chats.

Public follower totals miss the room behind the door

This creates a measurement problem.

Suppose a creator posts less often on a public network than five years ago. Their replies are down. Their visible comment section looks quieter.

That does not tell us whether their audience relationship weakened.

They may now spend that time talking to 300 patrons in a private chat, answering members on Discord, or posting material visible only to subscribers.

The opposite mistake is also possible. A large paid membership does not prove that everyone is actively participating. Membership counts, chat messages, views, and genuine relationships are different measurements.

But public metrics alone are incomplete once important interaction moves behind membership access.

The retreat can be economically sustainable

Patron spaces also solve a problem the old public web rarely solved well: they connect community participation directly to support for the person doing the work.

That can make a smaller audience more valuable than a huge passive one.

The tradeoff is visibility.

Search engines do not see the same conversation. Casual readers cannot browse it. Future researchers may find a creator’s public output and miss the community that existed around it.

For Dead Internet Theory, the lesson is straightforward.

A quieter public profile does not necessarily mean the creator lost the humans.

Sometimes the crowd became a club.

Posted on by

Paid newsletters and the movement of discussion behind subscriber access

A newsletter can look like a publication from the outside and behave like a private club on the inside.

That is increasingly normal.

Paid newsletter systems do more than deliver essays by email. They can also host comments, chat threads, direct messages, live discussion, and member-only spaces tied to the publication.

Substack’s current documentation makes the access model explicit. Publication owners can limit comments to free or paid subscribers, and Chat can be restricted so that only paid subscribers or founding members can view or reply. See Substack’s guide to comments and its Chat access documentation.

That means a writer can move not only the article but the conversation around the article behind subscriber access.

A bounded audience changes the discussion

For writers, the appeal is easy to understand.

A paying audience is smaller, more committed, and usually easier to moderate than the entire public internet. Members may have shared expectations about the subject. Drive-by abuse and spam become less common. The writer also has an economic reason to spend time with the community.

Readers gain something too.

A subscriber chat can provide direct access to the writer, other informed readers, private recommendations, follow-up discussion, and a sense that the conversation belongs to a recognizable group rather than an anonymous crowd.

Those advantages are real.

So is the visibility cost.

The article may be public while the interesting part is not

A nonmember might encounter a free post through search and assume they are seeing the public conversation around it.

They may not be.

The best correction, disagreement, anecdote, source recommendation, or follow-up explanation could be sitting in a subscriber-only thread.

Substack’s support pages explicitly note that subscribers without access to a paywalled Chat cannot view or reply to those threads. See How do I paywall my Chat on Substack?.

That knowledge still exists.

It is simply not part of the open searchable web in the same way a public blog comment section would be.

Paying for community is another kind of retreat

This is not evidence that newsletters are killing public discussion.

Writers can keep posts public. Comments can remain open. Paid communities can fund work that would otherwise never exist.

The Human Retreat question is about where the social layer goes.

If writers increasingly move their most engaged readers into subscriber-only discussion, then public observers see less of the conversation even while the community becomes more active and financially sustainable.

The audience did not disappear.

Somebody put the good table behind the velvet rope.