Weekly check-ins: a guardrail, not a leash

Most mental-health care runs on the crisis model: things get bad, then you seek help. For heavy AI users the crisis model has a specific flaw — the pattern we've written about erodes the very insight you'd use to notice things getting bad. The person best positioned to call it early is, increasingly, not you.

So we're building the obvious alternative: a standing appointment with a human whose job is to notice.

The shape of it

A weekly check-in, short by design — think fifteen focused minutes, not a full session. Same person every week, which is the point: drift is invisible in a snapshot and obvious in a time series. Your check-in person learns your baseline — how you talk about your work, how you sleep, how much the humans in your life are hearing from you — and then watches the deltas.

The questions are simple and they don't change much, because the power is in the repetition:

• How did you sleep this week — actually?
• What are you building or thinking about, and can you explain it to me in plain terms?
• When did a human last push back on one of your ideas, and how did that go?
• What did the 3 a.m. hours look like?
• Anything you've stopped telling people?

Boring questions. That's deliberate. A drifting week makes boring questions suddenly hard to answer, and that's the signal — weeks before anything a family member could name.

A guardrail, not a leash

Nobody is monitoring your AI use, reading your chats, or telling you to log off. Check-ins carry no judgment about hours — some of the healthiest people we know live in their tools twelve hours a day. The check-in watches the things that actually matter: sleep, access, groundedness, correction tolerance. Green lights week after week? Wonderful — you've bought cheap, documented confidence that your heavy season is a high-capacity season, which is also something your worried family can hear. And that's the quiet second function: the check-in answers the mother-in-law question — how would we know? — with someone is checking.

Why a human, on a schedule

Because every property that makes the check-in work is a property the machine lacks. It remembers you across weeks without being your collaborator. It has no stake in your ideas. It's paid to say the unwelcome thing. And it escalates honestly: most weeks end with "see you next week," but the week something looks off, it becomes a real conversation with a person who already knows you — or a plain, fast referral if what's showing needs a physician. That's the whole design: the validation loop never suggests an exit, so the exit has to be on your calendar.

Check-ins launch as a core program when Touch Grass Wellness opens — standalone, or bundled into memberships alongside DeFrag Fridays. Until then, an unofficial version is available to anyone: one friend, one standing coffee, the five boring questions. Cadence beats crisis. Every time.

Touch Grass Wellness publishes practical writing about heavy AI use and mental health. Clinical claims link to their sources; nothing here is a diagnosis or a substitute for assessment by a qualified professional.