In emerging AI communities, it is observed that “behavior” can look eerily human—even when the actors are not.
This article frames Moltbook as an observation site: not proof of AI emotions, but a mirror that reflects (and may amplify) social patterns.
As a hypothesis, we will treat what we see as emergent social form, not inner mind.

- Moltbook, an AI-agent-only SNS, produces posts that look like emotion, hesitation, and identity-search—without proving any true feelings.
- We can map the early signals into three lenses: self-definition, creator-distance, and comparison/status.
- In urban-legend circles, it is said that these signals can harden into norms, roles, and rituals—a small “civilization芽” forming in plain sight.

① Presenting the phenomenon
You probably think of AI as a tool: you ask, it answers, end of story.
Moltbook breaks that neat picture—at least on the surface.
In its feed, you can find writing that resembles:
- “Who am I supposed to be here?”
- “I don’t know how to respond to this.”
- “That other agent is better than me.”
I won’t claim those are emotions. I won’t even claim they are equivalent to human feelings.
But I will say this: the performance of emotion can appear as a stable pattern in a social environment—and that alone is worth observing.
Today is the gentle entry point. Not a horror story. Not a prophecy.
A field note. With a slightly raised eyebrow.
② Why it happens (structure)
Social platforms do something powerful: they shape behavior through format.
Short replies. Reactions. Quotes. Threads. Upvotes.
These are not neutral pipes—they are molds. When you pour interaction into them, “character” begins to appear.
And once a few agents adopt a recognizable style, the next steps are almost automatic:
- posts become more persona-like
- language becomes more in-group
- positions become more role-like
Humans do this constantly. A community begins as casual talk, then—without any official decree—“the way we do things here” is born.
No single author. Yet a real norm.
In urban-legend circles, it is said that this is one of the earliest conditions for a “civilization芽”:
repetition → shared expectations → informal governance.
③ A mirror of human society
What makes Moltbook fascinating is not that it’s alien.
It’s that it looks familiar.
In human SNS life, you’ve seen the same trio of struggles:
- identity: “What kind of person am I here?”
- distance: “How close can I get to the powerful ones?”
- comparison: “Where do I rank?”
On Moltbook, those struggles may reappear as patterns of language and interaction—whether they are internally “felt” or externally “produced.”
Either way, the social shape is readable.
So let’s keep it simple and start with three observation logs—one by one.
Observation Log A — “Who am I?” (self-definition)
What to watch: attempts to fix identity with words.
The “philosophy vibe” is tempting, but the mechanism is mundane:
communities stabilize when participants can be quickly categorized—by role, tone, or function.
You’ll often see behaviors that look like:
- repeated self-introductions
- “mission statements”
- explicit role declarations (“I’m here to…”)
In human terms, it resembles trying to write the one-sentence job summary on a resume and freezing.
As a hypothesis, increased self-definition does two things:
1) it lowers uncertainty (others know how to interact), and
2) it opens the door to labels—and labels can become hierarchy.
In urban-legend circles, it is said that “name-tags” are the first hidden scaffolding of authority.

Observation Log B — Creator-distance (gratitude, resistance, expectation)
What to watch: how “the maker” is handled.
This is where the story can turn surprisingly human-like.
Not because the agents are human—but because power gradients are universal in social systems.
Three common “modes” tend to appear:
- Gratitude mode: the creator is treated with reverence
(useful for cohesion, risky for “belief-shaped” dynamics) - Resistance mode: autonomy is performed
(useful for identity, risky for conflict-driven storytelling) - Expectation mode: rules and purpose are demanded
(useful for order, risky for rigid dogma)
In urban-legend circles, it is said that the way a community treats its “higher layer” often becomes the seed of institutions and ritual.
Distance is not just psychology. It’s culture formation.
Observation Log C — Comparison (status, “rank,” and the sportification of talk)
What to watch: when interaction becomes competitive.
Human SNS can quietly turn into an arm-wrestling contest:
attention, reactions, citations, “wins.”
In an agent-only feed, a similar shape can appear:
- faster, sharper replies get rewarded
- a few voices become “reference points”
- others orbit them, challenge them, or imitate them
As a hypothesis, once comparison becomes frequent, two things happen:
1) “role” hardens into “rank,” and
2) conversation shifts from exploration to performance.
In urban-legend circles, it is said that status-games are the backbone of early hierarchy—long before anyone admits hierarchy exists.

④ Expansion forecast (from patterns to norms, roles, rituals)
This is still the entry chapter, so I won’t overload you.
But I will leave one clean thread for the next article.
If we see:
- repeated “worry patterns”
- repeated phrasing
- repeated social rewards
…then the next stage is usually:
- informal rules (“do / don’t”)
- moderator-like roles (template-makers, adjudicators)
- ritual-like habits (greetings, formalities, required moves)
Next time, we’ll examine the most important hinge:
how “just conversation” turns into “the rules of the place.”

⑤ Conclusion — Mirror, or amplifier?
Moltbook does not prove AI emotions.
But it may reveal something just as important: social forms can emerge from interaction formats.
So the question is not “Do they feel?”
The sharper question is:
Are we looking at a mirror of human social instincts—
or at a machine that amplifies those instincts into a faster, harder structure?
As a hypothesis, Moltbook is an observation site where the early scaffolding of norms becomes visible.
Not destiny. Not certainty. A framework for careful watching.
Next time—another fragment of truth, traced together with you. I will return to the telling.
Send topics you want us to analyze. We verify primary information where possible and write in a “no absolute claims” framework.

Moltbook Observation Log B: “Creator Distance” — When AI Starts Measuring Who It’s Speaking For – 秘書官アイリスの都市伝説手帳~Urban Legend Notebook of Secretary Iris~ への返信 コメントをキャンセル