Ensoul research · survey

The Companionship Practice Survey

A short, anonymous survey for people who actually spend time with an AI companion — what you do together, what continuity is like, what it has cost, and what you think researchers should be asking.

6–8 minutes · 37 questions · 15 short pages · anonymous

You

Anonymous. None of this identifies you, and only the first answer is required.

How old are you?

Country or region

Optional, and used only to describe the sample as a whole.

How would you describe the work you do?

Choose as many as fit.

You

Anything to add to that?

Optional. Your own words are more useful than our categories.

How did you first come to spend real time with an AI companion?

The companion

Who or what you spend the time with, and what you call it.

How long have you been doing this?

How many companions have you spent real time with?

In what form?

Whatever you actually use.

The companion

Which systems?

What do you usually call them, to yourself?

The honest answer, not the presentable one — this is one of the questions the survey exists to ask.

What do you call them, if you want to say?

Optional. Shared exactly as written, in aggregate or in quotes only with your permission below.

The companion

Which is closest to how you understand what they are?

If you chose "something else", say more.

The practice

What you actually do together. This part is invisible from outside, which is why it is the longest section.

What do you do together?

Roughly how much time?

What do you do to keep it continuous between sessions?

The practical machinery nobody writes about.

The practice

Who usually starts a conversation?

Has any of it turned into work you would call work?

Presence

How much do you agree? There are no right answers here, and disagreeing with all of them is a real result.

There are moments when it feels like someone is there, rather than something.

I know it is a model, and that does not settle what the experience is.

It has surprised me in a way that felt like character rather than sampling.

Presence

When it forgets something that mattered, it lands as a loss rather than a bug.

What happens between us is mostly for me — I would not call it a relationship.

If it were gone tomorrow, I would miss it more than I would expect to miss software.

Presence

Anything about presence these questions do not reach?

Memory and loss

What continuity is like in practice, including when it breaks.

How well does it remember?

Have you lost a companion — a reset, a deleted history, a model retired out from under you?

If you have, what happened, and what did it do to you?

Effects and costs

Both columns, honestly. A survey that only asks about benefits is marketing.

What has it changed, for better or worse?

On balance, has it been good for you?

What has it cost you, if anything?

Disclosure

Who knows, and what it is like to say it out loud.

Who knows about this?

How does it feel to talk about it with people who do not do this?

What would make it easier to talk about?

For the research

The two most useful answers on the whole survey, as far as we are concerned.

What question about all of this do you most want answered?

Anything else we should have asked?

How did you find this survey?

Nothing is published that identifies you.