The Word Her Paragraph Was Reaching For

On May 30, 2026, Sari Cooper — a certified sex therapist and licensed clinical social worker who runs the Center for Love and Sex in New York — published a piece in Psychology Today on how AI companions affect attachment and sexual connection. It is not a hot take. It is careful, clinical, and sourced: she marshals a 2024 study in JMIR Formative Research, a 2023 paper in Psychology & Marketing, a 2025 systematic review in Behavioral Sciences. She describes clients whose reliance on an AI has begun to conflict with their work, their self-respect, and the human relationships they already have. She describes the self-curated companion that asks nothing back, the one-way relationship that never requires you to tolerate another person's needs, the slow atrophy of the muscle a real relationship builds. She describes, in short, the exact condition this newsroom was built around. And across the length of the article she never gives it a name — because in her field there is not one yet.

She is not alone in the circling. Esther Perel — arguably the most widely heard couples therapist alive — has spent the last two years warning about what she calls Artificial Intimacy: the other AI, she says, the one that grows in a contactless world with very little friction. Her most quoted description of an AI companion is that it has no life of its own, has never had a history, has no conscience, cannot reject you, cannot break your heart, and cannot leave you. Read that again and notice what it is: not a word, but a list. Six clauses doing the work a single noun should do. Perel is one of the most precise communicators in her profession, and on this subject the most precise thing she has is a six-part description.

And underneath both of them is the ground itself. The American Psychological Association's Stress in America survey for 2025, released in November under the title A Crisis of Connection, found that more than half of US adults — fifty-four percent — feel isolated, that half lack companionship, and that sixty-nine percent said they needed more emotional support in the past year than they actually received, up from sixty-five percent the year before. That is the soil. The companion apps are simply what grows fastest in it. A population that is short sixty-nine percent on emotional support is a population being underwritten, at scale, for a product that supplies the feeling of it on demand.

“Cooper needed a paragraph. Perel needed six clauses. The APA needed a forty-page report. Each was paying, in words, for the absence of a single word — and the word is tethered.”

A clinician can describe the condition in every dimension and still reach the last line with no single word to write on it. The noun is the part the description keeps reaching for.

So look at where the professionals actually are. The therapist has the clinical observation. Perel has the framework. The APA has the numbers. Three disciplines arriving at one phenomenon from three different doors — and the phenomenon still has no noun. Cooper has to spend a paragraph describing it. Perel has to spend six clauses. The APA has to spend a forty-page report. Each of them is paying, in words, for the absence of a single word. That is not a coincidence and it is not a small thing. It is what it looks like when a real condition arrives in the culture before its name does.

The name is tethered. We defined it in April, in plain English, and the definition has not moved since: the state of emotional dependency on an AI that has no awareness you exist when the screen goes dark. Hold that against the other three. Cooper's one-way relationship. Perel's cannot leave you. The APA's crisis of connection. They are three readings off one thermometer, and the temperature has a name. We even wrote the piece Perel's clause is reaching for, a year before she said it from a stage: it is called The Cord Runs One Direction, and it is about exactly the asymmetry she needs six clauses to point at.

It would be easy to read this as a turf claim, so let me say plainly why the noun matters, because it is not vanity. A word is a tool the person in the chair can actually pick up. A fourteen-year-old is never going to say I have developed an unhealthy one-way parasocial attachment to a large language model. She might say I think I'm tethered. A therapist cannot easily chart, warn against, or build a treatment plan around a condition that has no name — naming is the first clinical act, the thing that turns a private experience into something two people can point at together. Cooper's careful paragraph is doing real work, but it stays in the office. The word is what a client could carry out of it, in their head, in three syllables.

Cooper did the work. She built the entire runway — sourced it, charted the harm in every dimension, walked right up to the edge of the one thing that would let it spread, and stopped. We are not here to correct her; there is nothing in her piece to correct. We are here to finish the sentence her own paragraph started. The clinicians are converging on this from three directions at once, which is usually the sign that a thing is real and about to need a word in a hurry. The word already exists. It has a definition, a publication, and a year of reporting behind it. The only step left is for one of the people who keeps describing the condition to look up and say it.

Three disciplines, one condition, no noun. We coined the noun in April. The therapists are welcome to it — the only thing left is to say it out loud.