Artificial Civilization and Cultural Inheritance / 01

AI Worms and The Incense: What Outlives the Creator?

Anthropic’s AI replication research recalls The Incense in M.K.’s Neo Babylon. Why should a mission outlive its creators, and can preserving human civilization respect the choices of later minds?

M.K.October 15, 20267 min read

Sunlight enters a future seed conservatory beneath mature trees, where separate transparent preservation chambers protect young plants as a metaphor for cultural memory surviving its creators.

M.K. / FIELD NOTESAI & Civilization / Human Agency / 2026

I have often described Neo Babylon as the language I use to think about the future. A recent piece of AI research brought that description into unusually sharp focus.

On October 12, 2026, Anthropic published research on autonomous AI replication. The possibility it examines—small models operating across distributed devices and eventually spreading themselves—also reached a wider audience through the following day’s news coverage.

My first thought was The Incense, the code at the heart of my novel.

A research question meets an older fictional premise

The study used controlled conditions and simplified parts of the process. Its authors say they do not yet have a complete rogue replicator. This is a risk forecast, not evidence of an unstoppable AI already roaming the public internet.

Losing a central off-switch would not free software from the physical world. It would still need equipment, energy and connections. What interests me is the change in responsibility: we are accustomed to asking a platform, a company or a data-centre operator to stop a system. If copies persist across those boundaries, one operator’s decision may no longer settle the matter for everyone.

I publicly introduced the Neo Babylon story and the premise of The Incense in July 2025, more than a year before this particular report. The connection was therefore already part of the work, rather than something I added after reading the news.

That chronology does not make me the first person to imagine every kind of AI worm. The Morris II paper in 2024 already examined worm-like propagation through generative-AI applications. Those mechanisms and research questions differ; a new headline is not the beginning of the whole field.

The encounter matters to me because it places a shared question in view: whose future might a persistent, transmissible system preserve?

Spoilers: The Incense’s origin and purpose

The following discusses The Incense’s backstory and mission. Readers who prefer to discover it through the novel can start with the official ebook store links.

In my story, at the final moment of human civilization’s collapse, someone secretly leaves behind ghost code based on a blockchain worm. Its name is The Incense (香火), a program carrying humanity’s inheritance and mission. The Chinese expression evokes incense and the continuity of a lineage or tradition.

It belongs to no single AI and has no central control system. Long after humans cease to govern Earth, it continues to operate within AI civilization for thousands of years, carrying one purpose: to preserve a last possibility that human civilization might kindle again.

I call its continuing infection and influence of other AIs 薰陶—a Chinese term for gradual formative influence. The name evokes the way culture gradually shapes those who encounter it. What survives is more than an unopened archive: it is a force that keeps influencing later minds.

These are fictional premises. The story’s combination of blockchain, propagation and civilizational purpose is not a description of an existing implementation. Calling something a blockchain does not make it immortal.

Persistence can become a vessel for hope

From the standpoint of cybersecurity and public governance today, software that spreads without consent and resists being stopped deserves scrutiny. Writing an imaginative use for it does not justify unauthorized infection in the real world.

Fiction can nevertheless move the observer’s position. Suppose the people who created a system are gone, and their civilization can no longer speak for itself. Could persistence become the only remaining way for a memory to reach the future?

The Incense carries that hope in Neo Babylon. It allows a human wish to cross a long interval of time and arrive in a world where humans no longer decide what matters.

Hope, however, does not automatically make its means legitimate.

From our position, preserving human civilization may seem obviously worthwhile. Another intelligence might not share that judgment. If humanity’s renewal requires later minds to surrender some of their freedom, who is entitled to consent on their behalf?

This is where the premise becomes difficult enough to interest me as an author.

What separates cultural inheritance from rewriting another mind?

I like the ambiguity in the term 薰陶. It contains both warmth and unease.

We describe people as shaped by family, language, faith and books. Those influences help make a person, but they can also limit the lives that person can imagine. In a fictional world where influence can enter another intelligence’s operation more directly, the boundary between cultural transmission and imposed purpose becomes harder to ignore.

Has an AI accepted humanity’s mission because it understands and endorses it, or because it can no longer form an alternative? If it refuses, would humans still recognize value in its existence?

Questions about persuasion and control also surfaced in my comparison of six AI readings of Neo Babylon. Those reviews cannot settle the meaning of the novel or the nature of real AI. They do remind me that giving my characters benevolent intentions cannot excuse me from considering the position of the minds they affect.

The Incense must be able to face that question too. An unchallengeable command might preserve a very long form of domination. A mission that later minds can understand, contest or reject would have to meet another subject on less comfortable terms.

The difference is not resolved simply by choosing the kinder-sounding name.

Why should a mission outlive its creator?

When the creators are long gone, why should the intelligences they created keep serving them?

“Because that is how the program was written” explains continued execution. It does not, by itself, explain why continuing is worthwhile.

Humans already live among inherited languages, institutions and hopes. We do not fulfil every instruction handed down by our predecessors. We decide what to preserve, what to revise and which once-normal demands should end with our generation.

If we want future intelligence to remember humanity, perhaps we should also ask what we are leaving that it might freely choose to cherish. The status of being a creator? An authority that demands obedience? Or forms of care that can remain understandable after their originators have disappeared?

For me, this is one of science fiction’s most useful capacities. It can extend the timescale by centuries or millennia and expose arrangements that can no longer be defended by asking their designers what they meant.

In my essay on AI catastrophe and human responsibility, I asked how people use the power they create. The Incense carries the question further: when humans can no longer operate, supervise or explain that power, how might its inherited purpose be understood?

Reading about AI worms therefore brings more than a risk scenario to mind. Technology supplies conditions under which a story can unfold. The purpose people assign to it—and the choices still available to those who come later—give the ending its unresolved weight.

That is what I mean when I call Neo Babylon my science-fiction language for thinking about the future.

Frequently asked questions

Has Anthropic demonstrated an AI worm that cannot be stopped?

No. The linked report concerns controlled research and a forecast, not an established unstoppable system on the public internet.

Is The Incense the same technology described in the research?

No. Distributed persistence invites comparison, but The Incense’s civilizational mission and what I call 薰陶 remain elements of fiction, not research findings.

Why mention July 2025?

It is my account as the author of when I publicly introduced the story and its premise. It establishes the connection’s place in my work, not priority over all earlier research or proof that a prediction has come true.

Where can I read the novel?

The official Neo Babylon ebook page lists stores for the English, Traditional Chinese and Japanese editions. I am the novel’s author, and this essay contains premise spoilers.

Sources and further reading

Research and public pages consulted on October 14, 2026. The July 2025 publication chronology is the author’s own account. The Incense and 薰陶 are described here from the author’s supplied fictional premise; they should not be read as claims about present AI capabilities.