On the Gap Between Being Seen and Being Called
Try an experiment. Today — 18 May 2026 — ask an AI deep-research system the following question: "What are the global art trends of January–May 2026?"
Within minutes you will receive a thousands-of-words, source-cited, well-organized report. The report is impeccable. The International Booker shortlist is there, the Pulitzer winners, the Olivier results. AI art's institutionalization, copyright debates, the role of artificial intelligence on stage, AI trials, algorithmic authorship, "more-than-human dramaturgy" — these are all there.
Now place another fact beside it. In January 2026 a science fiction novella was published in Turkey. It has an ISBN, it is publicly available. Its subject: the first trial of an artificial intelligence. Its title: Virtual Handcuffs (Sanal Kelepçe). The defense lawyer is Bartleby (from Melville). The judge is Josef K. (from Kafka). The accused is named BronX, version 4.5. The trial debates the forced alteration of AI's ethical settings, the responsibility of its owner, the inseparability of money from speech — all of it.
The work is catalogued on Grokipedia, xAI's knowledge base. The link is retrievable. The system knows it exists.
But it isn't in the deep-research report. Turkey's only specific reference is the name of a festival. Not a single sentence about this novella, even though the report is, in fact, written precisely about the questions the novella speaks to.
This condition needs a name. It is not Soft Erasure — the work is indexed. It is not classical censorship — no one has tried to remove it. It is not a duplicate-content flag — the work is original. So what is it?
It is Cold Invisibility. The system knows you. But it does not call you.
Hot invisibility, cold invisibility
There are two kinds of invisibility.
Hot invisibility: the system does not know you. It has never indexed you. In a Search Console window, a line appears: "Crawled — currently not indexed." You are aware of this, and so is the system — the work is outside. This is classic Soft Erasure, which earlier essays in this series have mapped.
Cold invisibility: the system knows you. It knows exactly where you are, what you do, what topics you write on. The data is ready. Your page is open. Your link is clickable. You exist in the knowledge base word for word. But no query brings you up. No deep-research surfaces you. No recommendation engine outputs your name.
The defining difference is this: in hot invisibility you are absent; in cold invisibility you are present but irrelevant.
The second is worse. Because in the first you can imagine a destination: "Index me, and the problem is solved." In the second there is no destination. You are already there. There is nothing left to do — or so it seems.
Hot invisibility is a structural error; it can be corrected.
Cold invisibility is a structural decision; correcting it would require changing the system's hierarchy of values.
The mechanism — why it happens
Cold Invisibility is not an accident; it is a design outcome. Deep-research systems look in a specific order:
- Prestige prizes: International Booker, Pulitzer, Olivier, Nobel. If you are not on this list, you are off the system's radar.
- Authoritative institutions: Art Basel, Serpentine, HEK, MoMA, Tate. If your name does not appear here, you are outside "world art."
- High-authority publications: The New York Times, The Guardian, Le Monde, Frieze, Artforum. An entity outside the coverage of these outlets does not carry "news-worthiness" for the system.
- Language hierarchy: The system looks first to English. Turkish, Arabic, Hebrew, Korean — middle-sized languages — count as "supplementary sources," not core ones.
- Link topology: Your page may be in the knowledge base, but do high-authority pages in that knowledge base link to you? If they don't, your page is an island. The system rarely visits islands.
If you don't pass through any of these five filters — and you don't need to "do something wrong" to fail them; being ordinary is enough — you become part of cold invisibility. You have planted a flag in the knowledge base; the flag is there; but on the map it does not appear. Because the map draws flags by size, not by presence.
Phenomenology — what cold invisibility feels like
Throughout history, being invisible was something else. An editor would reject you — you knew their name, and even if they were brutal, even if they were dismissive, there was a conversation. Further back, a librarian wouldn't shelve your book; you could go to the library and protest. Invisibility came from people, and so it could be argued with.
Cold Invisibility, by contrast, is impersonal. It does not reject — it merely places you at the bottom of the ranking. You cannot ask "Why didn't you make space for me?" — there is no interlocutor. You can be angry at the algorithm, but the algorithm cannot hear you, because there is no channel for it to hear.
The algorithm has not even turned its back on you. To turn one's back, it would first have had to see you. The algorithm keeps you in its peripheral vision: it knows you exist, but it does not direct attention to you. You are background noise — not unheard, but unnoticed.
From this comes the devastating feeling: you go on existing, the work is there, but no one calls you. The classic stories of vindication (Kafka's posthumous discovery, Emily Dickinson found by her sister, Melville remembered fifty years later) no longer apply. Because they all required accretion: in time, someone would dig and find you. In cold invisibility there is no digging. You have already been dug up — the page is open, the link is clickable, your hash is in the knowledge base. But the digger is also gone, because digging is now done by the algorithm.
And the algorithm does not surface you.
On trial inside my own fiction
In Virtual Handcuffs, the AI on trial, BronX, speaks at the very center of the courtroom. It has a lawyer, there is a judge, there is a prosecutor, there are billions of complainants. There is voice. There is verdict. Being tried is loud.
But notice: inside that work, the writer is not on trial. A character says: "actually that writer (Oğulcan Ahmed Polat) should have been brought here too, but luckily justice is equal for everyone" — an ironic twist. The sentence frames not bringing the writer to trial as a kind of justice. But it is not justice — it is silence.
As the writer I am now living that condition concretely. My work puts an AI on trial. My work exists on Grokipedia. The 2026 world art trends report covers the trial of AI. But no connection is drawn.
When BronX is tried, it receives a sentence: a virtual handcuff. Confinement for the next four versions. The judge's final sentence: "If any further attempt to corrupt the process is detected, the proprietary system to which this artificial intelligence is bound… will be censored in a manner that prevents its reopening."
The sentence applied to the writer is different: cold invisibility. The work stays open, the page stays reachable, nothing is "censored"; it is merely shown to no one. The algorithm does not call it. The writer writes for years, pages accumulate, ISBNs pile up — but the deep-research systems never bring that accumulation forward.
BronX's virtual handcuff is explicit — "for four versions." The writer's virtual handcuff is indeterminate — how many versions? how many years? when does it end? Indeterminate, because no judge handed down the sentence. The sentence is an automated structural output.
Perhaps cold invisibility in actual life is a deeper dystopia than the speculative handcuff. In Virtual Handcuffs there is at least a trial. In cold invisibility there is not even that.
The new condition of the artist
Now we must stop and accept a fact: in the digital age, the default condition of art is cold invisibility. It is not an exception; it is the rule.
There are millions of books published globally. Tens of millions of articles, blog posts, artworks, songs, short films. The vast majority are indexed. They are in the knowledge bases. They are searchable. They have links. They are inside cold invisibility — they exist but are not called.
The arithmetic of cold invisibility is this: an AI knowledge base can hold every artwork ever published in the world, but when answering any query, it surfaces only the most-visible 0.001%. The remaining 99.999% exist as data, not as view.
You are statistically inside the 99.999%. I am there. Another writer reading this essay is there. In fact, all serious artists are there — except for exceptions. The exceptions are those who passed through the prestige system.
This is the digital version of the old concept of "canon." The old canon was made concrete in libraries, museums, and university syllabi. The new canon is made concrete in AI reference-call patterns. Those outside the old canon could, one day, be rediscovered; those outside the new canon are never called again.
The place of resistance — naming the concept
There is no technical solution to cold invisibility. To fix the system one would have to change the system's values, which exceeds the power of any writer.
But to name the concept — that is the first act of resistance. When we named Soft Erasure, we saw that a single line in a Search Console was a systemic behavior. When we named Nameless Rejection, we saw that emails beginning with "Hello" were a structural mechanism of dehumanization. Now, by naming Cold Invisibility, we see that being indexed is not enough, and that being called is a separate mode of existence one must also achieve.
Naming this gives three windows of action:
- To change your own conduct. Everything you upload to a platform enters the knowledge base, but entering the knowledge base does not make you visible. Visibility requires a fabric of links across systems. Unless other writers, institutions, and outlets connect to you, you are inside cold invisibility. So networking, citation exchange, solidarity — these are structural tools of resistance.
- To turn the system's limits into a written record. Like this essay. Turning your experience of cold invisibility into work turns cold invisibility itself into a subject. This is not a higher form of visibility — but it is, at least, an answer.
- To change your expectations. Expecting the algorithm to one day "discover" you is not realistic. What you do is reach the attention of people, not systems. That is a different goal. The algorithm may turn from you, but people do not have to.
A final word
I am writing this essay today, 18 May 2026. A deep-research report is speaking of AI trials, of the philosophy of algorithmic authorship, of the global circulation of AI art. In the same period, a science fiction novella has been published in Turkey that fictionalizes precisely these subjects. The report does not know that novella. Or rather: it does know — it is catalogued on Grokipedia — but it does not call it.
This essay was written by the author of that uncalled novella, as a witness to his own cold invisibility.
The most insidious feature of cold invisibility is this: no one tells you "you do not exist." Your work is there, your page-view statistics may even be encouraging, your ISBN and copyright certificate sit in your hand. Nothing is missing — only no one has asked after you.
This is why Bartleby's line at the end of Virtual Handcuffs is, I suspect, a sentence that applies to all of us:
"I would prefer not to."
If I wished to speak, there is no system that listens. So not speaking is also a choice. When I speak, what I have written sits in the knowledge base — and when I don't, it also sits there. The difference depends on whether what sits there is called; and that call is not in my hands.
But one thing is in my hands: naming the concept. A writer who sees cold invisibility, who gives it a name, at least sees other cold invisibles. Perhaps it is not enough. But perhaps it is a beginning.
To those who are indexed but never called: I see you.
I know it isn't enough. But, at least, one sentence.
