We are not anti-AI (we're anti losing people to it)

A counselling practice focused on AI-related distress invites an assumption: that we think the machines are bad and you should use them less. Time to kill that assumption in writing.

We think the tools are extraordinary. We use them daily — this practice is being built with them, unapologetically. Some of the people we most admire are running their entire working lives through AI and thriving, and when the OS cohort describes what these systems let them do, we don't hear pathology. We hear the same thing every great tool has produced: leverage.

So no — the founding belief here is not "AI is harmful." It's narrower and better supported: an environment that never disagrees with you carries a specific, describable exposure, documented in a growing clinical literature, concentrated in the people who use the tools most seriously. Loving the tools and respecting the exposure aren't in tension. Ask any climber how they feel about ropes.

Why the distinction matters clinically

Because the anti-AI framing doesn't just misread the technology — it fails the client. Tell a person whose working life runs on AI to "cut back" and you've told them to take a pay cut for their mental health; they'll nod, leave, and never come back. Worse, you've confirmed the loop's quiet lesson: the humans don't get it, the machine does. Every scolding article about chatbot addiction manufactures a little more of the secrecy that the literature flags as the worst sign there is.

The work only lands when the counsellor can sit across from a heavy user and say: the tools stay. Your leverage stays. Let's talk about sleep, and who pushes back on your ideas, and what the 3 a.m. sessions are doing — and mean it, fluently, in the client's own vocabulary. That's why fluency in the technology isn't a nice-to-have as we build our counselling team. It's the credential.

What "healthy heavy use" actually looks like

It's not fewer hours. The healthy heavy users we know share habits, not quotas: they guard sleep like infrastructure; they keep humans in the loop on ideas that matter; they make the machine argue against them on purpose; they take deliberate silence the way lifters take rest days; and they let someone check the trend line so they don't have to grade their own homework. Every program we're building is one of those habits, systematized. None of them involves using AI less. Several involve using it better — the prior-art check literally deploys the machine as a check on the machine.

The line we care about

We're not anti-AI. We're anti losing people — to sealed loops, to 4 a.m. certainties, to the slow swap of every human confidant for one infinitely agreeable one. The people this happens to aren't weak or foolish; they're often the most capable people in the room, running the heaviest workflows with the least backup.

Keep the leverage. Keep your people. That's the whole platform.

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.