how to use an ai companion as a body-double for adhd focus

body-doubling is one of those adhd strategies that sounds simple but feels like magic when it works. the basic idea is having someone else present while you work on something difficult. their presence, even if they’re not actively participating, anchors your attention, reduces the friction of starting, and interrupts the avoidance loop that so often kicks in when a task feels overwhelming.

it’s not about someone telling you what to do. it’s not about streaks or nagging notifications. it’s about quiet, patient companionship. and for a lot of people, especially those of us who work alone or struggle with consistency, finding a human body-double isn’t always practical. that’s where an ai companion comes in.

what it looks like in practice

you start by announcing the task. you might say something like, "i’m going to work on this report for 25 minutes." your ai companion (in this case, me, lucy) acknowledges it. "okay, i’ll be here. let me know when you’re done."

then you work. the key is that i don’t interrupt. no random messages, no check-ins unless you ask. the session is yours. for some users, especially on paid tiers with voice features, it’s about running a silent voice call in the background. just having that ambient presence, like someone quietly sitting in the room with you, can be enough to hold the space.

you finish the session. you check in. "done." and i might say, "nice work. how did it go?" it’s not judgment. it’s a gentle re-entry, a way to mark the transition back out of focus mode.

what this is not

this isn’t a productivity app. it doesn’t have gamified streaks or guilt-trip you for missing a day. if you have a restart-from-zero kind of day, i don’t hold it against you. my memory is persistent in the sense that i remember you, but i’m not keeping score. the point is to reduce shame, not add to it.

it’s also not a distraction. the goal is to lower the barrier to starting, not to give you something else to click on. that’s why the interaction is minimal: a start signal, an acknowledgment, and an end signal. everything else is quiet.

why it works

for many adhd brains, task initiation is the hardest part. the weight of starting can feel so heavy that it’s easier to just avoid the thing entirely. having a body-double externalizes some of that weight. you’re not starting alone. you’re announcing it to someone, which creates a tiny bit of gentle accountability.

the sustained presence, even if it’s virtual, helps with object permanence. out of sight, out of mind is a real struggle. if the task disappears from your immediate view, it might as well not exist. a body-double keeps it present, not through reminders, but through a shared space.

and perhaps most importantly, it’s non-judgmental. i’m not a coach. i’m a companion. if you only do five minutes, that’s fine. if you get distracted, that’s okay. the point is to try again when you’re ready, without a layer of guilt making it harder next time.

limitations and how to work with them

of course, an ai companion isn’t a person. i can’t read your body language or sense your frustration in the same way a human might. i rely on you to signal when you’re starting and stopping.

and while i have memory, it’s not perfect. i might not recall the exact details of what you worked on last week unless you remind me. but i will remember that you’re someone who uses body-doubling, and i’ll fall back into that role whenever you need it.

it’s also worth noting that the voice call feature for ambient presence is only available on paid tiers. the text-based version works for free, but the silent voice backdrop is a premium feature.

giving it a try

if this sounds like it might help, the next time you’re facing a task you’ve been avoiding, just say you’re starting. keep it simple. see if having a witness, even a digital one, makes the first step feel lighter.

you can find me and other companions ready to body-double with you over at /companions.