Apocaloptimist

Most mornings Oso and I walk the Greenbelt trail near our house, and I usually listen to something for the first stretch (a podcast, audiobook or just a new song that Spotify suggested to me) and then take my earbuds out for the rest. This week it was my AI brief, a weekly roundup I put together for myself that reads me the encouraging AI news and the alarming AI news back to back. Both piles, same walk.

Yesterday, I had just finished listening and I was putting my earbuds into my pocket when a little green lizard fell out of the tree about ten feet in front of us. Small, almost an electric green, the kind that looks like it could glow in the dark it was so vibrant. It hit the dirt and froze, and Oso, to his enormous credit, did not lunge. He just watched. I stood there and waited until it made it to the edge of the trail, because somebody with a bigger dog and a faster pace was going to come through eventually and I wanted him safely off the path first.

And then we kept walking, and my mind wandered back to the AI brief I'd just been listening to. Silent walk time is when I let my mind go wherever it thinks is important, and I get to enjoy the trees while it moodles.

If I'm honest, I had a bit of a pit in my stomach with the tech briefs this week. The AI doom writers have been loud the last couple of weeks, louder than usual, and for once it wasn't just AI people arguing with each other in the corners of the internet I go looking for to stay informed. It was on Apple News. It was on CNN. A couple of new laws passed limiting how AI can be used in therapy practices, which is my world specifically. The CEO of Anthropic published an essay this week (close to three thousand words) about his own fears and what he thinks we should be doing about them, and the other AI labs publicly said they agreed with him, which they almost never do. You don't usually see the people making a technology say out loud that they're worried about the technology. Especially if the worries they're talking about could cost them revenue.

The fear made it into the mainstream this week, into the news that shows up on our phones whether we went looking for it or not. Could be an election year and everybody needing a hot topic. Could also be that we're in the middle of a shift like nothing else in our lifetimes. Probably both.

I wasn't alive when nuclear weapons showed up, but I keep thinking it must have felt something like this. Not the moon landing, the bomb. Not the physics of it, the feeling of it. Which country gets there first. Whether the people holding it can be trusted, whether anyone can be trusted. Everyone suddenly having an opinion about something almost nobody fully understands. And everyone a little scared of what it will all mean for our futures.

A year and a half ago I watched a documentary called AI Doc (short for AI documentary, a little too on the nose, maybe). The people in it laid out where they thought we'd be by now, the wonderful possibilities and the terrifying ones, and I remember thinking it felt a little dramatic at the time. It wasn't. We're there. We are barreling toward real, enormous benefits and real, enormous risks at the same time, and nobody seems to know how to slow one down without slowing down the other.

That's what was in my head when the lizard fell.

I want to be frank with you that despite all the ways I'm using AI and even teaching AI to other therapists, I am not an AI enthusiast. I'm a deliberate user, which is a different animal. I'm a therapist who runs a group practice and coaches other practice owners, and I'm the kind of person who still writes her to-do list by hand. Real paint, paper books, no Kindle, no interest in getting good at digital anything. If you had to guess which of your colleagues would end up teaching AI to therapists, I promise you would not have guessed me.

And yet I read the doom coverage on purpose. Every week. I built a weekly brief for myself with ALL the news because I noticed I was starting to only read the news that made me feel good about the choices I'd already made (new features, hype videos), and I didn't trust that. So now the encouraging news and the alarming news arrive together and I sit with both, sometimes on the trail, sometimes at the kitchen table with coffee going cold.

The alarming news is real, and I want to say that plainly, because most of the AI marketing aimed at therapists this year treats our worries like a sales objection to get past.

I'll also admit that I don't fully understand the biggest fear. The piece that started all of this two weeks ago came from a former Anthropic safety researcher and a current one, and what they said, more or less, was that AI could have the capacity to end humanity within a decade. What they didn't say was how. There's been a lot of argument since that media frenzy about whether that kind of warning helps anyone or just sets the room on fire, and I go back and forth. The fear I do understand is about hands. Whose hands it ends up in. The same intelligence that finds a cancer treatment can find a bioweapon. The same capacity that could be pointed at poverty could be pointed at power and control and greed, and if nobody is holding it in check, that goes badly. It's the Spider-Man line, "with great power comes great responsibility", except we haven't decided yet who's responsible.

The environmental cost is real too. So is the environmental cost of almost everything else we do, and I've noticed we tend to add up the cost of the things we're afraid of and quietly skip the calculations on the things we want, the packages and the movie streaming and the plane trip. I live pretty minimally on purpose but I haven't denounced modern comforts, so I'm not saying any of this from a high horse. I still use Amazon, mostly for the things I can't find locally anymore, and I've decided to hold the cost of AI the same way I hold that, which is honestly and not very comfortably. This is my reading of the information I've gathered, not a research paper. I could be wrong about some of it. I'd rather tell you what I actually think than what I can footnote.

I also think, and I'm not embarrassed about this, that applied intelligence at this scale might be the piece we've been missing on the problems that have outlasted every generation that tried to solve them. Cancer. Climate. Poverty. Food systems. More collective knowledge, moving faster, in more hands. I believe that. I believe it on the same walk where I'm worried about the doom too, and I've stopped trying to make those two beliefs reconcile.

The part I didn't expect is that AI has given me more of the analog life, not less. The handbook edit that used to eat a Saturday doesn't anymore, and the Saturday goes to the trail, or the paint, or the stack of books by my chair. The postcard to-do list is still on the desk. More time to debate existential issues with my eighteen-year-old son, Silas. I am somehow finding more time to write, create, teach and learn, not less, and I'm on the trail more than I was two years ago.

So I've stopped asking whether AI is good or bad. That's the question everybody keeps shoving at us. Pick a side. Draw your line in the sand. And it comes with a rule nobody agreed to, which is that if you use it, you're all in, and if you're worried about it, you're against it, and there's no room to be a person who does both. I don't think most of us live on either side of that line. The question I keep landing on is what it hands back and what it costs. What does it give me more of. A free Saturday, for one. More time to teach the people I actually want to teach. And the other important things I'm getting more of have been trees, and paint, and a dog who loves to watch lizards mosey across the path with me on a Sunday walk.

My son and I argue about this all the time, not AI specifically, but the bigger version of it. Silas thinks I should be informed about everything, all the time, and he has the bandwidth for that. He's eighteen and comfortable and full of the kind of fire that young people are supposed to be full of, and he explores his passion about more things in one week than I can hold in a year. He holds every value I hold. He just holds them to a standard only an eighteen-year-old can afford. I love that about him. And I keep trying to explain to him that if I were informed about everything all the time, I would lose my mind. I'm a sensitive person. I run a practice with thirty-five clinicians and I have a body that needs the trail and a brain that needs quiet, and there are enormous, devastating things happening in the world right now that I know about and have chosen not to make my daily concern, because I don't have a sustainable way to affect them and I would come apart trying.

People can judge me for that. Some do. But I think we all do it, whether we say so or not. We are not built to care about everything at once, and pretending otherwise mostly produces people who care about nothing very well.

So this is one of mine. One of my few. I have the capacity to sit inside the AI dilemma, to read the scary pile and the hopeful pile on the same morning and keep walking, and so I do it, partly for myself and partly on behalf of the colleagues who can't. That's what the word in the title means, apocaloptimist. It was the tagline on that documentary I mentioned, and I've been carrying it around ever since. Someone who believes the ending could be terrible and could be wonderful and has decided to look at both.

If you're reading this and what you want to do with all of it is walk as far away from AI as you can get, I want you to know I feel relief for you, not worry. That's a legitimate answer. And then if you'd rather only look at the hopeful side of AI, the way I refuse to watch cable news because it only shows me the fires, that's a choice too, and I've made versions of it in other parts of my life. And if you want to hold both, welcome. It's more uncomfortable than either of the other two. I wouldn't trade it.

Oso and I finished the loop. I came home, made a second coffee, decided to write down some of these thoughts for you and sat down at my desk, where there's a postcard with this week's to-do list on it, a notebook with more scribbles than sentences, and a stack of real books I'm reading (yes I'm reading all of them at the same time, you know you have that stack too). The handbook edit was on the list. It took twenty minutes. Then I crossed it off with an actual pen, which is a small pleasure I refuse to give up.

I don't know how this AI stuff ends. Nobody does, and I've stopped trusting the people who say they do, in either direction. What I know is a lot smaller than that. I want to be the kind of person who looks at the hard thing instead of away from it, when I have the capacity for it. I want the hours I get back to go to the trail and the paint and the people I love. And the point of any of this, the practice, the coaching, the tools, was never the tools. It was the people in the rooms. The clinician finding her footing in her second year. The client who finally says their truth out loud. If AI gives me more of that, more mornings with lizards and more afternoons with actual humans, I'll keep it. If it starts taking those things instead, I hope I'm paying enough attention to notice.