Most serious questions about platforms begin with something small enough to look trivial. In this case, it was a line: the nearly blank page that appears when I open someone’s WeChat Moments and find no visible posts.
At first, I treated this as a narrow question about blocking. Why does discovering that another person has hidden their Moments from me sometimes feel hurtful? Why do some users respond by immediately hiding their own Moments in return? Is this merely an individual sensitivity, or has WeChat created a small but consequential social mechanism through which people classify, test and sometimes retaliate against one another?
The question became more difficult as the discussion continued. A privacy setting appeared to be doing several things at once. It controlled access to posts, signalled the perceived distance of a relationship, altered expectations of reciprocity and supplied ambiguous evidence from which one person might judge another. The interface did not determine the judgment, but it created the conditions under which the judgment became reasonable.
I did not arrive at that formulation immediately. The inquiry moved through several explanations that were plausible but insufficient. The AI helping me first overestimated the ambiguity of the blank page. I corrected it using my observation of the actual interface. It then separated privacy rights from relational respect. I accepted part of that distinction but objected that it made the social meaning of exclusion too easy to dismiss. It later argued that time-limited visibility was technically different from a middle position between openness and blocking. Again, the technical point was correct, but my objection revealed a prospective dimension of recognition that the technical analysis had missed.
The same pattern recurred when we turned to personal websites. The AI proposed the familiar model of a website as the canonical home and social platforms as distribution channels. I rejected routine distribution through Moments because my articles are long, frequently written in English and addressed to readers defined by intellectual relevance rather than by their presence in my contact list. That objection produced another distinction: publishing something publicly is different from automatically delivering it to everyone one knows.
What began as a question about a blank line consequently became an inquiry into relational legibility, bounded freedom, platform classification and different forms of publicness. This second account preserves how those ideas emerged, including the explanations that had to be corrected before the larger problem became visible.
The initial asymmetry
My own practice shaped the original question. I rarely prevent anyone from seeing my Moments, apart from accounts dominated by advertising. At the same time, I have not posted there for years. My profile therefore communicates inactivity rather than targeted exclusion. It may indicate that there are no posts from the last thirty days. That distinction matters to me because another person can see that my silence is general. I have not created a wall specifically around them.
I do not expect unrestricted access in every relationship. If somebody adds me from a university group containing hundreds of members and immediately selects “Chat Only,” I do not regard that as offensive. We may be contacts for one limited purpose. Neither of us has represented the connection as a friendship.
The situation that disturbed me was more specific. Suppose another person adds me first, begins a direct conversation and asks me for information. At first I can see their Moments. Later I cannot. They may still be able to see mine, and ordinary messaging remains available. The interface now presents a relationship in which one person retains access and the other does not.
My first reaction was that this could easily create conflict. Another user might respond according to an “eye for an eye” principle and hide their own Moments. Even if I did not retaliate, I might revise my judgment of the person. The sequence could make me wonder whether I had been treated primarily as a tool: useful enough to answer a question, but insufficiently recognised to remain within the person’s social audience.
I was cautious about converting that feeling into a conclusion. The other person might have many reasons. They could be reorganising contacts, protecting themselves after a bad experience, separating professional and personal life, or applying the same restriction to a large category of acquaintances. Yet the possibility of an innocent explanation did not remove the friction. It only made the source of the friction harder to identify.
I eventually realised that the final setting was not enough to explain the experience. The route to the setting mattered. I began thinking in terms of a relational delta: the difference between the relationship as it had been presented and the relationship after the visibility change.
A blank page encountered on the first day of a weak tie may carry little meaning. The same page appearing after initial openness, a request for assistance and continued access in the opposite direction can communicate a demotion. The important object is therefore not the isolated permission but the transition:
initial contact → visible self-presentation → request or exchange → unilateral withdrawal of visibility.
This did not prove instrumental intent. It made an instrumental interpretation plausible. That difference between evidence and motive remained important throughout the inquiry.
What the Zhihu discussions changed
I initially wondered whether people had simply lost the desire to share. Moments seemed quieter than before, and the blank profiles appeared to support a broad story of declining expression. The first collection of Zhihu answers complicated that explanation.
People described avoiding Moments because almost every kind of disclosure had acquired a social cost. Posting success could be interpreted as showing off. Travel, academic achievement, money, relationships and family happiness might stimulate comparison, envy or requests for assistance. Posting unhappiness could expose weakness, produce gossip or create material that might later be used against the person. Even ordinary daily sharing could be condemned as boring, excessive or performative.
The emerging logic was severe. If I show that I am doing well, someone may resent me. If I show that I am doing badly, someone may judge me. If I share frequently, I may become irritating. If I share nothing, I become mysterious or socially absent. Silence begins to look like the safest strategy.
One response to a person explaining why they no longer posted on Moments captured the contradiction with a joke: after refusing to disclose on WeChat, the person had told the whole story on Zhihu.
That joke changed my model. The desire to share had not necessarily disappeared. It had migrated.
People who seemed silent on WeChat might be writing extensively on Zhihu, Xiaohongshu, Douban, Bilibili, short-video platforms or pseudonymous forums. The expression had moved from an audience organised by offline relationships to audiences organised by topics, interests or temporary attention. The quantity of disclosure could even increase while the person’s Moments remained empty.
The second Zhihu discussion—why somebody would hide their Moments without deleting a contact—added another layer. The answers did not reveal one shared etiquette. They revealed several rival understandings of what a WeChat contact is.
Some users treated privacy control as a sovereign right. Exchanging WeChat details was compared to exchanging business cards; being listed as a contact created no claim to personal disclosure. Others treated Moments visibility as evidence of friendship. If a familiar person excluded them, they would delete the contact because the relationship no longer had substance. Some defended quiet restriction as an act of civility: they wanted to maintain workable contact without creating the confrontation of deletion. Others described keeping people because future coordination or exchange might still be useful.
Several comments stated the retaliatory logic explicitly: if the other person could see my life while preventing me from seeing theirs, the arrangement felt unfair, so reciprocal blocking restored balance. This supported my initial observation, but it also showed that the conflict did not arise from blocking alone. It arose from competing moral models of reciprocity.
Under one model, each person controls their own space and owes nothing to the other. Under another, mutual contact creates an expectation of approximately reciprocal visibility. Under a third, the relationship can remain civil and useful without becoming personal. WeChat uses the word “friend” for all of them.
A later Zhihu discussion about the apparent decline of Moments introduced the workplace more forcefully. Respondents described WeChat as a mixture of family communication, employment, advertising, institutional promotion and private life. Some reported needing to consider supervisors, colleagues, clients, relatives and acquaintances before publishing a single post. Others described organisational demands to repost official material or provide evidence that the repost remained visible.
At this stage, the original question about individual mentality became too narrow. The problem was no longer simply why a particular person had hidden their posts. It was how a communication platform had accumulated so many incompatible relationships that users were forced to repair the resulting context collapse through private permission settings.
The first neat explanation was wrong
When I asked what a blank or nearly blank Moments page actually meant, the AI offered a seemingly comprehensive list:
The person may never have posted; older posts may be outside the visibility window; posts may have been deleted or made private; everyone may have been excluded; or only you may have been excluded.
The answer was cautious and therefore initially sounded credible. It warned against overinterpreting the line. Yet it did not fit the interface behaviour I had observed.
I objected that, in the situations and app versions familiar to me, a person who had never posted might have no Moments entrance on the profile at all. A general time limit such as three days, one month or six months was commonly marked as a temporal policy. If everyone had been excluded, an external observer could not discover that fact. In a case where ordinary messaging remained possible and the person had previously displayed posts, a relationship-specific restriction was a more natural interpretation than the unrestricted list of possibilities suggested by the AI.
This was a productive error. The AI’s original list preserved an important caution: the line does not disclose one unique cause. What had to be abandoned was the stronger implication that all explanations were equally plausible in the concrete situation.
WeChat’s own explanation later supported a more precise formulation. In 2024, the company responded to claims that long and short versions of the line revealed different relationship states. It explained that the visual length reflected software versions. In newer versions, deletion, blacklisting, “Chat Only” and exclusion from Moments could all produce the same short line (WeChat Team, 2024).
The official explanation did not make the line meaningless. It showed that several socially distinct actions had been compressed into the same output. If messaging continued normally, deletion or blacklisting became less likely, leaving a smaller range of access-related explanations. The result was neither certainty nor total ignorance. It was structured ambiguity.
This changed the diagnosis. The problem was not that the recipient knew nothing. The recipient knew enough to suspect a relational restriction but not enough to identify its reason or scope. The platform protected the privacy of the person setting the boundary by transferring interpretive labour to the person encountering it.
The episode also changed how I used AI in the inquiry. AI-generated comprehensiveness could conceal an incorrect weighting of possibilities. My observation of the interface supplied evidence that the abstract taxonomy lacked. The revised account became possible only after I rejected an answer that was verbally polished but empirically under-calibrated.
The language of rights did not settle the relationship
The next important disagreement concerned respect. The AI proposed:
Access to someone’s private posts is not itself a right. Respect and equal visibility are not identical.
I accepted the first sentence provisionally. Nobody owes me a complete archive of their private life. Yet I doubted that the second sentence adequately described the relationships I had observed. When a person keeps me permanently outside their Moments while maintaining access to mine, it often seems difficult for the relationship to develop genuine mutual respect offline. This remained a first-person observation rather than a universal law, but it required more than a restatement of privacy autonomy.
The AI’s distinction was useful because it prevented me from treating visibility as an entitlement. It was insufficient because it answered the legal or moral-permission question without answering the relational question.
I eventually separated three levels that had been compressed together. Minimal moral respect requires acknowledging another person’s autonomy and refraining from coercing disclosure. Relational respect concerns whether two people treat each other as participants of reasonably equal standing. Intimacy goes further and usually involves mutual vulnerability, responsiveness and some form of self-disclosure.
A person may be morally entitled to hide their Moments and still communicate a relationally ungenerous position. Rights determine whether the person may set the boundary. They do not determine what the boundary means within a particular history.
Privacy answers, “May I withhold this?” The relationship asks, “What does this withholding mean between us?”
This formulation also allowed me to preserve exceptions. Two people can disclose very different amounts while respecting each other. One may never post but speak openly in private conversation. Another may maintain a consistent rule that excludes every colleague while remaining generous and trustworthy through direct interaction. Equal standing does not require identical content.
The stronger and more careful conclusion became:
Respect does not require equal disclosure, but it normally requires symmetrical standing or another credible form of reciprocity.
My grievance is therefore strongest when several conditions converge: the other person initiated contact, requested attention or assistance, retained access to my self-presentation, withdrew their own visibility and offered no alternative form of mutuality. Under those conditions, the feeling of being used is not proof of motive, but neither is it an irrational reaction to a neutral interface.
Research on reciprocal self-disclosure helped contextualise the intuition. Comparable disclosure can make a person feel understood, validated and cared for (Reese and Orrach, 2023). Experimental research on computer-mediated communication also associates turn-taking disclosure with greater liking and trust (Chen et al., 2024). These studies do not establish a duty to provide Moments access. They explain why unilateral opacity can affect how a relationship is experienced.
Helen Nissenbaum’s account of contextual integrity supplied another useful correction. Privacy is not adequately described as the total quantity of information hidden from others. It concerns whether information flows according to the norms of a particular context (Nissenbaum, 2010). My concern was therefore not simply “I want to see more.” It was that the performed relationship and the informational relationship could diverge. Someone might approach me as a cooperative peer while configuring the platform to treat me as a purely functional contact.
I still could not infer one mentality from the setting. The same button might express safety, work–life separation, caution after harassment, a general policy towards weak ties, effort minimisation, low trust, temporary discomfort or instrumental interest. This produced a further question: if one setting can enact so many intentions, why does the platform display all of them through the same relational surface?
Why time-limited visibility changed the model
The next revision arose from a more technical distinction. The AI argued that “visible for three days” was not really a midpoint between openness and blocking. Blocking concerns who belongs to the audience. A three-day or one-month limit concerns how much of the past remains visible. These are different axes.
I agreed with the technical classification but objected to the social conclusion. Time-limited visibility performed at least three functions that made it a real relational middle ground.
First, it indicated that I had not been individually excluded. Second, it left me within the audience for future posts. Nothing might be published tomorrow, but if the person did publish something, I could still encounter it. Third, the restriction expressed a general concern about historical exposure rather than treating me as a specially designated threat or “Chat Only” contact.
The technical distinction survived, but its meaning changed. Temporal visibility and audience inclusion are different controls. A temporal limit can nevertheless mediate socially between complete historical openness and relational exclusion.
The decisive insight concerned the future. I had initially thought of visibility as access to an existing archive. The discussion revealed that it also concerns membership in a possible future audience.
A time limit says, “I restrict the depth of my visible biography, but I have not removed you from the future audience of my life.”
A person whose profile currently contains no recent posts may still recognise me prospectively. The social information lies in my continued eligibility to see what might be posted later. I began calling this prospective recognition.
This distinction explained why two equally empty pages could feel different. A labelled three-day window limits the amount of history available to me. A relationship-specific bare line changes my status as a viewer. One regulates retrospective depth; the other can regulate relational membership.
Research on post visibility in WeChat supports treating these configurations as part of relationship management. Interviews with young Moments users suggest that people use visibility, following and unfollowing to evaluate, rescale and sometimes quietly terminate relationships. Expectations of reciprocity help determine which mediated connections remain meaningful (Hu and Zhan, 2026).
The study does not prove that all viewers interpret a temporal window as I do. It does support the broader claim that visibility is woven into the maintenance of networked intimacy. The setting is not merely a storage-retention parameter once it becomes visible within a relationship.
From Tencent’s algorithm to the human algorithm
My original formulation asked whether Tencent had intentionally designed an algorithm that forced users to construct “my circle” and “your circle.” That question contained two ideas that later needed separating: algorithmic recommendation and algorithmic classification.
Moments has traditionally emphasised chronological ordering rather than the recommendation logic associated with short-video platforms. It would therefore be misleading to claim that a recommendation algorithm directly decides which friends belong to my intimate circle.
Yet the absence of recommendation does not produce a neutral social space. Every profile remains the output of an access-control computation. In simplified form:
visible posts = published posts ∩ time window ∩ audience rule ∩ relationship state
What I see is not another person’s single public Moments page. It is a viewer-relative projection generated from their posts and the rules attached to our connection. If V(A,B,t) represents what A can see of B at a given time, there is no requirement that V(A,B,t) equal V(B,A,t).
This changed my image of Moments. The platform is called a “circle,” but technically and socially it resembles a matrix of pair-specific projections. Each user appears differently to different contacts. Two people connected through the same interface may inhabit radically different versions of their supposedly shared social space.
The platform delegates much of the classification work to users. People tag, filter, exclude, time-limit, unfollow, counter-exclude and delete one another. The machine does not have to infer the relationship because it asks human beings to encode it. Users become operators of what I began to call the human algorithm.
This algorithm does not consist only of source code. It is a feedback loop:
interface category → user classification → visible relational signal → interpretation by another user → changed behaviour → new classification.
One person hides their Moments. Another notices the line and hides theirs. The first person encounters the changed visibility and interprets it as retaliation. A privacy setting has become an input to the next social action.
I also began to see an ontological compression at work. A real relationship may contain uncertainty, professional dependence, gratitude, caution, curiosity, temporary conflict and unequal degrees of intimacy. WeChat asks the user to express this complexity through a small set of states: visible, hidden, “Chat Only,” time-limited, unfollowed, deleted or blocked.
Once selected, these categories can appear to reveal what the relationship already was. In reality, the interface has helped produce the relationship by giving one interpretation a durable technical form.
This reframed the question of Tencent’s intention. I found no evidence that WeChat’s designers wanted to create hostility or encourage retaliation. The settings answer genuine needs: contact-list overload, professional exposure, harassment, advertising and the desire to preserve civil communication without granting intimate access.
At the same time, the resulting conflict cannot be dismissed as entirely accidental or external to design. The options were intentionally built, even if the interpersonal injuries were not intended outcomes. The more accurate conclusion is that WeChat designed the conditions of classification, while users supplied motives and judgments within those conditions.
There is evidence that WeChat’s leadership recognised the structural problem. In 2019, Zhang Xiaolong said that, given another opportunity, he would separate the personal album from the Moments feed. He acknowledged that the arrival of work contacts, business contacts and sellers had made users increasingly hesitant to express ordinary emotions (Zhang, 2019).
This suggested path dependence rather than a deliberate plan to destroy intimacy. Moments originated under an assumption that contacts could plausibly be treated as friends. WeChat later became payment infrastructure, workplace communication, an address book of weak ties and an entry point for commercial and public services. Privacy switches became patches applied to a social graph that had outgrown the category of friendship.
Why the choices feel natural and compulsory
This raised another question from my original inquiry: why do users accept these choices so readily? Why do people debate whether another individual is good or bad without asking why the platform represents the relationship in this way?
Part of the answer is that the choices are experienced as private and local. I change a setting for my own comfort. The interface frames the action as personal privacy management, so its effect on the other person appears secondary. The cumulative social system disappears behind a sequence of individual decisions.
The consequences are also distributed. The person selecting “Don’t Let Them See My Moments” may never witness the recipient’s confusion, hurt or retaliatory decision. Each user encounters only one side of the permission. The emotional cost can be externalised to someone who is absent from the configuration screen.
Repeated use then naturalises the categories. “Chat Only” begins to appear as an objective kind of person rather than a product-defined access state. Three-day visibility begins to function as a personality type. The blank line becomes evidence of hostility, secrecy or social rank. Platform categories migrate into ordinary descriptions of character.
Network effects further limit the available freedom. Users can theoretically refuse Moments, maintain multiple accounts, move to another platform, use private groups or publish on a website. Yet WeChat’s role in everyday communication makes complete departure costly. Formal choice remains broad, while practical choice narrows.
I therefore stopped saying that users have literally “no choice.” That overstates the technological determination. A more accurate concept is bounded agency. Users make real decisions, but they decide within a taxonomy they did not design, inside an infrastructure they may be unable to leave, under social expectations produced partly by earlier users of the same system.
When the friend list becomes infrastructure
The workplace material made this boundedness more concrete. WeChat combines family, friendship, employment, institutional communication, payment and commercial contact. The heterogeneous audience creates what social-media research calls context collapse.
Goffman’s account of social presentation helps explain why this is tiring. People ordinarily show different aspects of themselves in different settings. Audience separation is part of ordinary social competence rather than proof that one identity is false (Goffman, 1959).
A study based on interviews with fifty-one urban WeChat users distinguished between adapting to collapsed contexts and trying to restore them. Some users reduced expression to content safe for everyone. Others rebuilt boundaries through filtering, multiple spaces, role management and temporal limits. Participants described expanding contact lists and work transitions as making relational filtering increasingly messy and effortful (Li et al., 2024).
This distinction clarified a pattern in the Zhihu material. Some users adapted by posting only achievements, neutral photographs, work promotions or material that could survive every audience. Others attempted restoration by hiding posts from colleagues, using multiple accounts or moving different forms of expression to different platforms. When the labour of restoration became too high, silence was the remaining adaptation.
Research on how users “re-domesticate” WeChat after it becomes disruptive similarly shows that people try to recover boundaries between private life, work and public obligation (Huang and Miao, 2021). The platform becomes infrastructure first and social space second.
The problem is even harder because WeChat permissions are primarily dyadic while social life is often triadic. I may hide a post from a supervisor but leave it visible to a colleague who can take a screenshot, mention it or display it in conversation. I can correctly configure the software and still fail to contain the information socially.
Permissions control who can look directly. They cannot control what a network does with what one person has seen.
This helps explain why temporal windows and individual filters may feel ineffective. The user is not only calculating the risk of each authorised viewer. They are calculating the unknown paths through which information can travel after the first viewing.
Chronological ordering also proved less neutral than it initially seemed. A seller, employer or institution that publishes repeatedly can dominate the feed without recommendation algorithms. Posting frequency becomes visibility power. As ordinary personal expression decreases, promotional material constitutes a larger share of what remains. The altered content gives users still less reason to browse, reinforcing further withdrawal.
The system can consequently remain statistically active while becoming phenomenologically hollow. Millions may still open Moments every day, while the specific experience of reciprocal personal disclosure weakens within many networks.
The stranger-privacy paradox
Once I stopped assuming that people had lost their desire to share, a new contrast became visible. WeChat is technically restricted but socially consequential. Platforms such as Zhihu or Xiaohongshu may be technically public but socially less entangled with the user’s offline life.
A mixed-method study of personal-experience sharing found that participants were comparatively willing to disclose to real-name acquaintances or to strangers under anonymity (Xu and Zhang, 2025). The result helped explain why a person might say very little on Moments and write extensively for unknown readers elsewhere.
I began thinking of this as the stranger-privacy paradox. The unknown reader may learn more about my inner life than my colleague does, but the stranger has less practical power to affect my daily relationships. An acquaintance may see only a carefully controlled profile while possessing access to my workplace, family or reputation.
Privacy is therefore partly about consequence rather than audience size. A small audience of identifiable people may feel more dangerous than a vast audience of strangers. Disclosure migrates towards spaces where negative reactions are comparatively disposable.
This produces an inversion of intimacy. Strangers encounter thoughts, doubts and long-form reflections. Acquaintances encounter the socially safe shell. The self is not disappearing; it is being distributed across platforms according to different risk structures.
Research on cross-platform behaviour supports the idea that users maintain different presentations according to audience, norms and purpose (Davidson and Joinson, 2021). The different versions are not necessarily false. They are contextual aspects of a person who cannot speak appropriately to every audience at once.
What the evidence could and could not establish
The Zhihu discussions were important to the development of the argument, but they could not establish how common each behaviour was. The questions attracted users who had already noticed blocking, jealousy, workplace pressure or declining activity. Dramatic stories of betrayal and retaliation were more likely to receive attention than uneventful accounts of consistent privacy management.
I therefore treated the answers as a qualitative map of moral interpretations rather than a representative sample. They showed what blocking could mean to users, not how frequently every meaning occurred.
A similar problem arose with statistical claims circulating in the later discussion. Numbers attributed to “QuestMobile 2025 data” included large decreases in Moments posting and interaction, widespread use of three-day visibility and high rates of inactivity among young users. These figures initially seemed capable of confirming the story of decline.
When I looked for the underlying public report, however, I could not locate those specific feature-level figures or their methodology in the report to which they were attributed (QuestMobile Research Institute, 2026). The absence of a traceable table, sample definition or measurement procedure changed how much weight I could place on them. They remained circulating claims rather than confirmed evidence.
Official figures create a different interpretive problem. In 2019, Zhang Xiaolong reported that approximately 750 million people entered Moments daily (Zhang, 2019). Such scale demonstrates that Moments was far from literally dead. It does not tell us whether personal posting became concentrated among fewer users, whether commercial content occupied a growing share, whether particular life stages were withdrawing, or whether visiting remained habitual after reciprocal expression declined.
The evidence therefore required a more discriminating research design. A serious study would need to document current interface behaviour across app versions and relationship states. It should follow users through transitions such as entering employment, changing workplaces or accumulating clients. It should distinguish personal posts, advertisements, compulsory reposting, browsing, liking and direct conversation. It should examine relationship sequences before and after visibility changes rather than taking a single screenshot as self-explanatory.
It would also need to avoid unethical attempts to expose private settings. The purpose should be to understand how people interpret relational transitions, not to create surveillance tools for discovering who has blocked whom.
A particularly important question remains whether the apparent difference between younger and older users reflects generation or life course. University students may still post frequently because their contacts remain socially coherent. The same users may become cautious after several years of employment. Without longitudinal evidence, youthful activity cannot prove that Moments will remain personally expressive for that cohort.
The personal website enters the argument
My personal website first appeared in the conversation as an exceptional practice. I write long essays, often in English, and I prefer to compose and read them in the sustained environment of a computer rather than inside a mobile feed. Very few of the people in my WeChat contact list are likely to read such work regularly.
The website also requires technical labour. I need to manage a domain, hosting, a CMS, design, updates, backups and security. I can customise the architecture and retain a durable archive, but that autonomy depends on skills and continuing maintenance that most users may not want. My practice therefore cannot stand as a representative solution to the decline of Moments.
Initially, the AI described the personal website as a “third model” that could reunify different slices of identity into a durable authorial presence while requiring readers to approach voluntarily. I found this useful because it distinguished the website from both the collapsed acquaintance audience of WeChat and the fragmented pseudonymous identities of other platforms.
The website can be professional and personal without placing every form of expression into one feed. Its architecture gives context. A homepage may function as a foyer, a biography as an office, technical writing as a workshop, reflective essays as a library and photography as a gallery. The sections belong to one authorial identity, but visitors choose which rooms to enter.
Personal homepages have long combined autobiography, expertise, self-presentation and creative work (Dominick, 1999). What seemed “old-style” about the personal website began to look philosophically significant. It organises publication around artefacts and places rather than around a compulsory stream addressed to contacts.
In Moments, the initiating question is:
Which people in my contact list may see this?
On a website, the question becomes:
What am I prepared to make public, and how should a reader encounter it?
The site can therefore perform what I think of as architectural context restoration. Different dimensions of the self remain related without being flattened into one message for one collapsed audience.
Its lack of automatic distribution is both a weakness and an ethical advantage. Friends may never notice that I have published something. Search engines may not surface it quickly. Comments and feedback may be scarce. Yet the article does not interrupt people merely because they once exchanged contact information with me.
A page can remain available until a reader has a relevant question. This creates a form of slow discoverability. Search, citation or a carefully chosen link may bring the reader months or years later, when the subject rather than the social relationship makes the encounter meaningful.
Why the first website model was still insufficient
The AI then proposed a familiar model:
The website as the canonical home; platforms as optional distribution edges.
This resembles the IndieWeb practice of publishing on one’s own site and syndicating elsewhere. The canonical copy remains under the author’s control, while platforms provide visibility and social interaction. Research on IndieWeb systems demonstrates both the resilience of owning the original and the continuing technical precarity created by dependence on corporate APIs (Jamieson et al., 2022).
I accepted the value of the canonical home but rejected routine syndication to Moments. That objection changed the model again.
My articles are usually too long for the kind of attention I associate with a mobile social feed. Many are written in English, while most of my WeChat communication occurs in Chinese. Their relevant audience is defined by subject knowledge or research interest rather than by friendship, family connection or historical acquaintance.
Posting every article to Moments would therefore create several mismatches at once: language, genre, device, attention and audience. It would distribute the work widely among contacts while reaching relatively few people prepared to engage with it.
The response signals would also be ambiguous. A like might indicate that somebody read and agreed with the article. It might instead express personal support, politeness or acknowledgement of the relationship. The social signal could become more visible than the intellectual encounter.
I therefore arrived at a more selective model:
The website is the canonical home, and distribution follows relevance rather than contact.
If an article concerns another person’s academic or professional field, I can send the link directly and explain why I thought it might interest them. This resembles scholarly correspondence more than social broadcasting. It still requires care. The relevance should be genuine, and the message should not manufacture an obligation to read, respond or praise.
This approach separates friendship from readership. A friend does not have to prove friendship by reading a long English article outside their interests. A serious reader does not have to become my WeChat friend in order to encounter my work.
The absence of automatic updates is therefore no longer only a disadvantage. It creates a consent boundary between availability and delivery. I began calling this non-intrusive publicness.
Something can be publicly available without being automatically addressed to everyone I know.
This distinction also revised my understanding of sharing. People may not object to making something public. They may object to converting publication into compulsory delivery to a socially consequential audience.
The personal website does not solve every problem. Public pages can be scraped, copied, misunderstood outside their original context and rediscovered years later. Search engines influence visibility. Hosting providers, registrars and software remain infrastructural dependencies. Maintaining the site requires continuous labour, and independence can produce isolation as easily as autonomy.
Nevertheless, the site establishes a different relationship between author, work and audience. Moments begins with the social graph and asks what content may flow through it. The website begins with the work and allows an audience to form around it.
How the question changed
Looking back, I can reconstruct several transitions that the polished conclusion alone would conceal.
I began by wondering whether blocking was a minor etiquette problem. The Zhihu discussions showed that it belonged to a wider migration of disclosure away from accountable acquaintance networks.
I initially treated the blank line as fairly direct evidence of personal exclusion. The AI warned that several states could produce it. I then corrected the AI’s overly broad list using the interface behaviour I had observed. The resulting concept was structured ambiguity rather than certainty or complete ignorance.
I initially suspected that a person who hid their Moments might be acting disrespectfully. The AI replied that privacy access was not a right. My objection forced a distinction between moral autonomy, relational standing and intimacy. The revised position preserved privacy while taking asymmetry seriously.
I initially understood three-day visibility as an imperfect privacy compromise. The AI separated time from audience. My response revealed that temporal restriction also preserves future audience membership. This produced the concept of prospective recognition.
I initially asked whether Tencent’s algorithm was dividing people. The chronological nature of Moments challenged the recommendation-algorithm formulation. Access-control logic and retaliatory feedback then revealed a different mechanism: the platform delegates classification to users, who become operators of a human algorithm.
I initially treated my website as a niche technical alternative. The discussion reframed it as another form of publicness. The AI’s syndication model was useful but still too platform-oriented for my practice. Rejecting routine Moments promotion led to the principle of distribution by relevance and the distinction between public availability and automatic social address.
At each stage, part of the earlier explanation survived. The blank line remained ambiguous, but not infinitely so. Privacy remained a right, but not a complete account of relational meaning. The time window remained technically distinct from audience filtering, but became a social middle ground. Platforms remained useful distribution tools, but distribution itself became optional rather than the assumed purpose of publication.
What remains unresolved
I now think the phenomenon is worth studying, but the central object is broader than blocking. It is relational legibility: how a platform allows people to know, infer or misinterpret their standing in one another’s lives.
WeChat gives users genuine powers of privacy and boundary management. It also represents several different relationships through almost identical outputs. A person protecting themselves from harassment, a worker separating colleagues from family, a friend quietly withdrawing after conflict and someone maintaining a purely instrumental contact may all produce the same line.
Greater transparency might reduce friction, but it would create new problems. If WeChat indicated that a restriction was general rather than individual, recipients might feel reassured, but the platform would reveal more about the user’s privacy policy. If it displayed reciprocal visibility, users might monitor and retaliate more intensely. If it required explicit labels such as “colleague,” “acquaintance” or “friend,” those categories could harden social hierarchies instead of clarifying them.
Design cannot abolish the pain of every boundary. Sometimes exclusion is the purpose of the boundary. Yet good design could avoid making too many socially distinct acts collapse into one unexplained symbol. It could more clearly separate an address book from an intimate audience, distinguish a personal archive from a social feed, and make general temporal policies legible without exposing specific exclusions.
My own judgment remains provisional. Nobody owes me unrestricted access to their private posts. Yet freedom to control privacy does not make selective invisibility socially neutral. If someone initiates contact, asks for my attention or assistance, retains access to my self-presentation and then withdraws their own visibility, I may reasonably revise my understanding of the relationship. I should not claim to know the person’s private motive, but I do not have to pretend that the configuration communicates nothing.
The three-day window taught me that an empty page can still preserve prospective recognition. The blank line taught me that ambiguity can govern a relationship. The Zhihu material taught me that sharing may migrate rather than disappear. The personal website taught me that I can publish more openly while broadcasting less.
The “circle” produced by WeChat is therefore neither entirely mine nor entirely yours. It is a negotiated boundary generated from two people’s decisions inside a system neither person designed. The platform does not dictate the final moral judgment, but it supplies the categories through which we make one another visible, invisible and judgeable.
The inquiry ends with a better question than the one with which it began. Instead of asking only why a particular person blocked me, I now ask:
What kind of relationship becomes possible when a platform makes access to one another both privately configurable and publicly interpretable?
References
Chen, Q., Zhang, Q., Zhao, S., and Li, C. (2024). “Trusting Strangers: The Benefits of Reciprocal Self-Disclosure During Online Computer-Mediated Communication and Mediating Role of Interpersonal Liking.” International Journal of Psychology, 59(1), 143–154. https://doi.org/10.1002/ijop.12957.
Davidson, B. I., and Joinson, A. N. (2021). “Shape Shifting Across Social Media.” Social Media + Society, 7(1). https://doi.org/10.1177/2056305121990632.
Dominick, J. R. (1999). “Who Do You Think You Are? Personal Home Pages and Self-Presentation on the World Wide Web.” Journalism & Mass Communication Quarterly, 76(4), 646–658. https://doi.org/10.1177/107769909907600403.
Goffman, E. (1959). The Presentation of Self in Everyday Life. Doubleday.
Hu, J. M., and Zhan, E. S. (2026). “Networked Intimacy via Post Visibility: Why Do People Unfollow WeChat Friends?” Communication Quarterly. https://doi.org/10.1080/01463373.2026.2643642.
Huang, Y., and Miao, W. (2021). “Re-Domesticating Social Media When It Becomes Disruptive: Evidence from China’s ‘Super App’ WeChat.” Mobile Media & Communication, 9(2), 177–194. https://doi.org/10.1177/2050157920940765.
Jamieson, J., Yamashita, N., and McEwen, R. (2022). “Bridging the Open Web and APIs: Alternative Social Media Alongside the Corporate Web.” Social Media + Society, 8(1). https://doi.org/10.1177/20563051221077032.
Li, P., Cho, H., Shen, C., and Kong, H. (2024). “From Context Adaptation to Context Restoration: Strategies, Motivations, and Decision Rules of Managing Context Collapse on WeChat.” Journal of Computer-Mediated Communication, 29(1), zmad043. https://doi.org/10.1093/jcmc/zmad043.
WeChat Team. (2024). Explanation of the Long and Short Lines Shown on WeChat Profiles, reported by National Business Daily. https://m.nbd.com.cn/articles/2024-03-23/3292739.html.
Nissenbaum, H. (2010). Privacy in Context: Technology, Policy, and the Integrity of Social Life. Stanford University Press.
QuestMobile Research Institute. (2026). 2025 China Mobile Internet Annual Report. https://www.questmobile.cn/research/report/2031215896219979777/.
Reese, Z. A., and Orrach, K. (2023). “Reciprocal Self-Disclosure: Although Respondents Are Reluctant to Steal the Spotlight, Self-Disclosers Feel Validated, Understood, and Cared for When Respondents Share Comparable Experiences.” Journal of Social and Personal Relationships, 40(11), 3485–3514. https://doi.org/10.1177/02654075231174530.
Xu, H., and Zhang, P. (2025). “The Influence of Anonymity and Social Ties on Personal Experience Sharing: A Comprehensive Mixed-Methods Study.” Proceedings of the ACM on Human-Computer Interaction, 9(1), Article GROUP31, 1–22. https://doi.org/10.1145/3701210.
Zhang, X. (2019). Remarks on Redesigning WeChat Moments, reported by People’s Daily Online. https://it.people.com.cn/n1/2019/0110/c1009-30515444.html.