Body Doubling for Autistic Adults

Body doubling is discussed most in ADHD spaces, but many autistic adults use it too — usually needing a different setup, with less social load and more predictability.

What the work is actually like

Task initiation and switching difficulties are common in autistic adults, and so is the cost of unpredictable social contact. That combination makes standard body doubling awkward: the mechanism helps, and the usual delivery — a stranger on video, small talk, an unfamiliar format each time — carries a load that can outweigh it.

Where presence helps

  • Starting, and switching between tasks
  • Externalising structure so it does not have to be generated internally
  • Predictable, repeated sessions that become routine

Where it does not — and may make things worse

This is the part most articles skip. Presence raises arousal, and raised arousal pushes you toward what you already know how to do. That is excellent for well-learned work and actively unhelpful when the job is to find something you do not yet have.

  • Formats with unpredictable social demands — new person each time, cameras, open chat
  • Anything where managing the interaction becomes the main task

A session that fits this work

  • Prioritise predictability: same time, same format, ideally the same person or none at all
  • Camera off wherever possible — presence does not require a face
  • Agree the shape explicitly up front, so nothing about the session is improvised
  • Consider an AI companion or a recorded ambient room, which are fully predictable by construction

The failure mode for autistic adults

Assuming a format that helps ADHD peers will help you. The scheduled-stranger-on-video model is popular precisely because unpredictability creates pressure, and pressure is not what you were looking for.

If none of this is landing, the problem may be upstream: read what body doubling actually is, or why starting is the hard part.

Common questions

Is body doubling helpful for autistic people?
Many autistic adults report that it is, particularly for starting and switching. What tends to differ is the delivery: less social improvisation, more predictability, and often no camera.
What if being watched makes it worse?
Then use a format with no observer: an ambient room, a companion that notices task-level activity rather than a person watching you, or simply someone in the next room. The useful part is presence, not being looked at.

Body doubling, elsewhere