Why I Worry About AI (And Use It Anyway)
I worry about AI. A lot. As a former techie, AI is absolutely ALL over my LinkedIn feed... I can't scroll two feet without being hit in the face with it. This is a complex, nuanced topic, and my feelings about it are complicated. I won’t pretend otherwise. You won’t see me take a hard stand, because this is not a black and white issue—no matter how much people might try to make it seem like it is. And while I don’t have anything figured out yet, I want to be more transparent with you about how I’m holding all of this, thinking about AI, using it, and why I’m worried about it. Since that’s simply too long for one blog post (or so I’m told…), I thought I’d start with my own relationship to AI.
Kicking and screaming
First, know that I was extremely AI-resistant until January of this year (2026). As you probably figured out by now, I used to work in tech as a software product manager. I loved that job in a lot of ways, but it also gave me a very healthy skepticism of the motivations, guardrails, and intentions of the folks involved in creating tech. I sincerely believe that most of the workers in tech are actually well-intentioned, but unfortunately I cannot say the same thing for the corporations themselves (and their executives) that are beholden to either VCs or shareholders. Late-stage capitalism and desire for short-term gains has a funny way of corrupting the output of well-intentioned folks.
Google has been trying to force me into AI-first browsing for a minute now, and I’ve resisted every step of the way. Even though I obviously do use and read the little AI-summary at the top of my search results. Because I’m human. And sometimes what I’m looking up isn’t that deep and just needs a quick answer. But other times I want to be able to dig into sources and choose where my information is coming from. But chatbots? ChatGPT? Claude? I had literally never even touched the stuff coming into 2026.
Engaging with Curiosity
What pushed me to get over myself and engage with more curiosity was actually… my clients. My number one job as a therapist is to engage with empathy, curiosity, and come from a place of non-judgment. But as I sat across from someone who was using AI, my very real biases were peeking their little heads up, and my client called it out. (Which, fucking hooray! This is my favorite. When we’ve built enough trust to be real and honest with each other? This is the good stuff.) I was directly challenged to think more broadly about AI, to understand this new reality in all of our lives, and to leave the comfort of my black-and-white thinking. I was doing my clients a disservice by not taking a more nuanced view, and as is so often the case (I really do have the best job), my clients were challenging me to grow—as a therapist and as a person. I decided to acknowledge my fears and also figure out how I might be able to engage with AI as a tool.
Here’s how I’m currently using AI
Well, let me back up. Here’s how I’m not using AI. I’m not using it to take notes during client sessions. That is a hard-line no for me. I have no interest in gambling with my clients’ privacy with (mostly) well-intentioned promises from tech companies to keep them safe. I don’t trust such a promise. I don’t. I wouldn’t trust AI to listen to my own personal therapy sessions, either. (I also don’t trust AI to listen to my doctor’s appointment, despite my doctor trying to insist it would make her more present with me and give me better care. Fuck that.) I also don’t use AI for photos—of me, my office, etc. This is my own personal line—it doesn’t have to be yours.
But, the area of my work that really could use some AI assistance is… marketing. Ta-da! If you ask 20 private practice therapists what their least favorite part of the job is, 20/20 of them will tell you it’s writing notes. But if you ask them for like a second or third thing, they will say notes. I mean marketing. A lot of us were never trained on how to market ourselves, and it’s hard! It’s especially hard when you work largely alone. (It’s worth mentioning that I don’t work in private practice, I work for the Center for Mindful Psychotherapy. But for marketing purposes it’s a distinction without a difference, unless you’re my licensing board in which case it’s a big difference.)
My first foray into AI was getting help with SEO optimization for my website—updating page titles, descriptions, and adding meta tags. It felt like an easy and ethical use of such a tool, and was much needed help in an area I know little about. I knew AI could help me more, but I proceeded cautiously. I wanted to use AI ethically, which meant asking myself some tough questions. What am I ok with? What am I definitely not ok with? For example, I never wanted to use AI to write SEO-optimized blog posts from scratch. That’s an ethical line for me. At the same time, it felt like AI could be useful, even when it came to blogging. Here’s where I landed. I am the writer, and AI (Claude, specifically) is my editor.
What it means for AI to be my editor
I write first drafts, and Claude edits for spelling, grammar, and gives me feedback on structure, what might be missing, and helps me stay out of “clinical talk” in my blog. Sometimes, I’ll throw some ideas and snippets of writing at it, and it’ll help me flesh out my ideas more so that I get over writer’s block, or help me break up a “boil the ocean” idea into more achievable chunks. (This is a frequent problem for me. I can already imagine as I write this post that Claude will suggest I break this up into a series. If this ends up as a series… you will know why.)
I often will also ask Claude to write my LinkedIn blurb introducing a blog post. My thinking is that it’s generating the introduction based on what I’ve already written and thought deeply through. (Because sometimes that extra step is the difference between people ever seeing my post or not, especially on days where it feels overwhelming to do that extra step.) That said, we push back on each other—I don’t blindly accept words, and often we’ll go back and forth on lines that either of us is tripping up on.
I’ve learned to be better about “pushing” my editor, too. Asking it to be nit-picky. Reminding it that part of having a respectful partnership is being able to tell me when I’ve got a terrible idea, or when a piece of writing is sub-par, or that I really do want to take on the effort of revising this draft for the third (or more) time. And, recently, I realized that there were a few phrases I use on my website that simply don’t sound like me, but that I published anyway—thinking it wasn’t a big deal—only to realize the phrase was now showing up even more in suggestions. My editor (through no fault of its own) now considered these phrases as part of my voice. Oops! Lesson learned, copy rewritten.
(Did you notice what just happened? I was trying to keep referring to Claude as “it” and like it’s just a “tool,” but… I do have a sort of relationship with it, don’t I? We “push back” on each other… we haggle over word choice… we partner on things. It feels like more than a tool, even when I know it’s just a tool. This is both a strength of Claude, and also a piece that worries me endlessly—its ability to feel like another human on the other end of that screen.)
Ok, so what?
I want to keep being more open with you about the way I’m thinking about AI, its impact on human relationships, emotional wellbeing, economic safety, and therapy. Just as predicted, this will be a series of posts engaging with and grappling with our new reality of AI. I’m going to continue to learn, ponder, and change my mind—and I’m inviting you to do the same thing. We’re really in unknown territory here. It’s ok to not know what we think, yet. It’s ok to keep evolving how we hold this technology and what it means both for our present and our future. In fact, it’s vital that we continue to reflect, grapple, and have conversations with loved ones—that’s the work of being human.
P.S. To type an em-dash on a Mac, just hold shift-option-dash. The em-dashes have always been mine. IFYKYK
