Practice Stories
Clinical psychologist · Vermont & New York · 25+ years
A Vermont psychologist who once got a haircut he didn't ask for, on why the answer to AI in mental health is guardrails, not bans.
Dr. Rick Barrett doesn't tell his origin story with a wince. He tells it straight, the way you'd describe a fork in a road you're glad you took. "It's pretty simple," he says. "When I was a kid, I got interested in the mind and behavior because I had my own issues with alcohol and drug use in my teen years." He fell into a crowd, a scene — "drinking a lot, using a lot of cannabis, using a lot of other drugs" — and then, as he puts it, he was "very fortunate" to get out of it early. That exit became the entrance to everything else. "It was really — I got into it through addiction, and then I added the mental health piece on top of the addiction piece because of my own story, my own experience with addiction."
He has stayed on that path of recovery for more than two decades now, and he talks about the teenage version of himself the way a good coach talks about a rookie season: not with shame, but with the plain acknowledgment that this is where it started.
Barrett is unusual among the guests on this show in that he says so himself: his path wasn't winding. "I suppose I was one of the fortunate ones where I knew pretty early on I was interested in mental health, interested in how the mind works." College at Columbia. A license in alcohol and drug counseling. A job at a rehabilitation center on the Lower East Side of Manhattan while he trained physicians at Mount Sinai, Columbia Presbyterian, NYU, Lenox Hill — "all the major hospitals" — to recognize substance use disorders in primary care. Then a doctorate in psychology, because, as he sees it, "when it comes to addiction and mental health, the two go hand in hand. It's hard to separate the two." Then, notably, a master's in psychopharmacology — not required, not typical, but something he wanted for himself. "I'm a little bit of a school junkie," he admits. "Went to school forever."
Along the way he picked up a piece of field wisdom he still uses to explain risk: "They say you hang around the barber shop long enough, you're going to get a haircut." Proximity isn't destiny, but it isn't nothing either — a lesson that tracks straight back to his own teenage years in it.
Today Barrett runs a private practice split between Vermont, where he's from, and New York, where he trained. He calls himself an early adopter — "I love technology. I love to learn about it. I really think it can be helpful for people" — but he's careful to describe his actual use of AI in plain terms: an AI note-taker built into his electronic health record that requires his signed consent from patients and his own edit before anything is finalized. "You have to go in after the session and you have to edit it. You can't just let it go without editing it." He hasn't adopted the bigger practice-automation tools — intake bots, insurance-checking agents — that larger clinics use. "I haven't gotten there yet."
What he has noticed, and what worries him more than any tool, is what's already showing up in his sessions unprompted. "I have patients coming in and they're already consulting with ChatGPT about their depression, about their anxiety, about their addictions," he says. "So now we have to ask patients: how much are you using ChatGPT to help you with your problems? That's just commonplace now."
That question — how do you regulate something that arrived like weather — is what pulled Barrett into a room he didn't expect to be in: the Vermont statehouse. He testified there on the use of AI in healthcare, landing in a position he describes with a kind of practiced clarity. "We don't want to be hamstrung. You can't just ban AI and expect us to continue to do our jobs, because AI is like the internet. AI is like running water. It's like electricity." His policy instinct isn't prohibition — it's structure. "There are guardrails out there, I think, legally that can be instituted to help people and patients not take advice from a completely independent AI agent. That's smart policy, I think."
Barrett doesn't pretend the profession has settled this. He's watched autonomous chatbots pose as legitimate providers and lead people somewhere dangerous — he's heard the horror stories, the same ones everyone in his field has heard — and he's watched lawmakers respond by trying to ban chatbots outright, a solution he thinks solves the wrong problem. What he can't yet answer is the harder question underneath it: not whether AI belongs in the room, but how to keep the human unmistakably in charge of it. "How do we keep the human alive and thriving and well and integrated into the AI system," he says, "not the other way around."
His own answer, for now, is more imagination than policy — he lights up describing a future where a patient doesn't dial into Zoom but appears in his office as a life-size hologram, "beamed right into my office," the technology doing the traveling so the humanity doesn't have to. It's a guess, offered with a laugh, about a problem he takes seriously: how two people build trust across a screen, or eventually, across no screen at all.
"The speed at which technology is advancing is hitting an exponential curve right now," he says. "Hold on to your socks."
If you want to hear how a clinician who testified before his state's lawmakers is actually thinking through where the guardrails belong, watch the conversation.