In the summer of 2025, I started reading about artificial intelligence seriously, and I have not been able to stop.
It began with Leopold Aschenbrenner’s Situational Awareness, a long essay a former OpenAI researcher published in 2024, arguing from the steady growth in computing power that human-level AI was plausible by 2027 and superintelligence before the decade was out. One line from it has stayed with me ever since:
Everyone is now talking about AI, but few have the faintest glimmer of what is about to hit them.
From there I went to the sources themselves. The essays the lab leaders publish under their own names. The system cards released with each new model, where the companies describe, in careful language, what their newest systems can now do and what worried them in testing. The safety research, and the forecasts written by people who had left the labs and were willing to say plainly what they expected. Month by month, I watched capabilities that had been predictions arrive as products.
What I was reading was not hype. It was a convergence of capabilities arriving faster, and with deeper implications, than almost anyone outside a small circle had registered.
Chicken Little
What I felt was alarm, and for a while I felt like Chicken Little, which is an uncomfortable thing for a psychologist to notice about himself, since alarm is one of the states I am trained to treat rather than to occupy.
The alarm was really about a gap. On one side was what I was reading. On the other was ordinary life around me: colleagues who had never opened ChatGPT, friends who thought of it as a curiosity, a tech person’s concern, something that would sort itself out.
That gap is what sent me to the desert and the ocean. I spent much of that year going out to Joshua Tree, where there is no signal, and to the Southern California coast to surf, because I needed somewhere quiet enough to remember what I actually believe about the human person before I said anything about what was at stake.
You cannot prompt a wave. The ocean asks one question, which is whether you are actually here.
I never resolved the tension between that alarm and everything else I feel about this technology. I have stopped trying to.
I am not standing outside this
I like these tools. I use them daily. I run a small fleet of them on my own work, and they have made me faster at things I am glad to be faster at and given me access to thinking I could not have reached alone.
I also believe we are living through a change in kind rather than in degree. Intelligence is about to be metered and delivered the way electricity was, and likely more cheaply and more completely than that. When a civilization gets a new utility it does not simply gain a convenience. It reorganizes around it. Cities move. Work changes shape. What counts as a skill changes, and so does what counts as a person’s contribution.
That is a paradigm shift, and it is not first a technology story. It is an anthropology story, which is why a psychologist ended up writing it.
What the book actually asks
Humanity in the Age of AI is not a book about what we will do with artificial intelligence. It is a book about what it will do to us.
Its questions are old ones arriving under new pressure. What is a person, and what is intelligence? What is memory, and what is freedom? What is work actually for, and what happens to a human life when the productive part of it can be generated on request? What must not be outsourced, and what safely can be?
Underneath all of it sits one claim. AI is not creating our crisis. It is revealing one that was already here, in our attention, our memory, our desire, our capacity to be present.
The machine is not the subject. It is the light, and what it illuminates is us.
The deepest danger was never that machines would begin to think. It is that we would quietly stop, handing over our attention, our judgment, and eventually our willing, one convenience at a time, until we reach the end of an ordinary life having authored almost none of it.
The turn I did not expect
I went out there braced to write a book about what to protect. I came back with something harder.
Artificial intelligence does not threaten human meaning. It clarifies it.
That is a more demanding claim than the threat narrative, not a gentler one. When the machine takes the drafting and the analyzing and the routine cognitive labor that educated people have used to understand their own worth, it does not destroy our capacity for meaning. It strips away the meaning we had assembled out of being productive, and leaves whatever we were actually made for standing in the open.
The person who loses their work to a system and asks what they are now for is asking the oldest question there is. The tradition’s answer to it is not consoling. It is vocationally demanding. You are for love, for real encounter, for the care of persons in their vulnerability and the accompaniment of persons in their suffering, for the moral deliberation that requires someone willing to be responsible for the outcome.
How I think God moves in a moment like this
Here is the part I hold most seriously and can least prove.
I do not think the future is secured by better algorithms. I think it is made of persons, formed or unformed, present or absent, rooted or adrift. And I think that in every age that has accelerated past what people could absorb, God has quietly kept a people who did not accelerate with it.
They are never the impressive ones. A widow praying her Rosary in 1952 does not know she is forming the faith of a grandchild who will form the faith of a great-grandchild who is ordained in 2031. She just prays her Rosary. That hiddenness is not a defect in faithful witness. It is its normal form.
There will be people in this decade who move fast, adapt constantly, and integrate every new capability the week it ships, and they will be productive and probably influential. There will also be people who move more deliberately, who protect what the environment is eroding, and who carry something through the acceleration that the acceleration cannot produce. The second group will not look significant. They rarely do at the time. But they will be transmitting what it means to be human, and they will be available for the work this decade is about to need most: the care that requires presence, the teaching that requires formation, the accompaniment that requires an interior life.
Every civilization is finally shaped by the things it stops believing are worth practicing.
I wrote this book for the people who intend to keep practicing.
It is out now
Humanity in the Age of AI: What It Means to Be Human When Machines Think is available in paperback, hardcover, and Kindle.
Paperback · Hardcover · Kindle
If you read it and it says something you needed said, a review helps more than almost anything else I could ask for.
If this gave words to something you have been carrying, share it with someone who needs it.
The Inner Exodus is free, and I intend to keep it that way. If these essays feed you, you can become a patron. A paid subscription buys nothing extra. It simply keeps the door open for the next reader. Thank you for walking this with me.
From the year behind this book
Turning Forty in the Wilderness
This is much closer to what interests me about AI than the endless discussion of which model performs which task better. If machines eventually become extraordinarily capable, the question doesn't become less human—it becomes more human. What is work for? What is intelligence for? What makes a life meaningful when productivity is no longer our primary measure of usefulness? AI may force us to revisit questions that philosophy has been asking for thousands of years.
Tobin is right that AI strips away the meaning we had assembled out of being productive, but 'clarifies' is too gentle a word for what comes next. A chrysalis does not clarify the caterpillar; it dissolves it, and I made the harder case for what that means here: https://firepit.substack.com/p/humanity-is-in-a-chrysalis