A World Where Things Just Work
Returning to Union Station
This isn’t an essay about how to build or use AI systems. It’s about what it feels like to live with them, and what that tells me about the kind of work I want to do.
The Union Station universe by E. M. Foner is where I can breathe. There are more than twenty‑five books in the series, and I’ve read all of them more than once. I found the series during a period of professional change that felt challenging, when the ground under my work was shifting in ways I didn’t yet understand. Reading it made things feel stable.
I revisited the series later, during a time of sustained stress across both my personal and professional life. I follow the author, so I know whenever a new book comes out and consume it immediately upon release. Not because I need something new, but because returning to that universe helps me relax. It helps me believe things will be okay.
What’s interesting to me is not the technology in these books, but how my body responds to the world they describe.
In the Union Station universe, the AI is competent. So competent that most of the time, things just work. Life is not perfect. People still struggle and make mistakes. But there is an underlying trust that the systems will support them. Not everyone uses AI. Not everyone trusts it. Some people keep their distance or choose other ways of living. Others accept AI, and aliens, and all the strange breadth of that world as part of a larger, wonderful universe. Difference is normal. Support is assumed. No one is constantly bracing for collapse.
In the real world, AI systems do not feel like that. They are powerful, impressive, and increasingly capable, but they ask something of us all the time. Attention. Supervision. Judgment. They need to be checked, validated, monitored, tuned. Trust is conditional and often provisional. Even when systems are working well, there is a background hum of vigilance, a sense that someone needs to be watching, ready to step in when the edges show.
That is not a failure of the work. It is the reality of building complex, probabilistic systems in the open world. But it means that relief does not come easily. Stability is something we actively create and continuously maintain.
A world where things just work feels lovely. Where the sink breaks and a robot fixes it. Where you need a small nudge to move forward and it arrives in a supportive way. I’m not saying I want to give up control over my life. I just don’t want to have to double‑check everything, all the time.
In the real world, AI still needs supervision. It doesn’t quietly decide you need a chocolate bar and deliver one to your door. But imagine a world where it could. Imagine how rested we might feel if some things were handled because someone else already saw the value in doing so.
Lately, I’ve been thinking about this in terms of how we build and use AI. A one‑off prompt demands constant attention. A reusable instruction still asks for oversight. A skill carries a little trust. An agent carries more. Each step isn’t about intelligence. It’s about how much of the work I can stop holding.
What I want, I think, is not sentience. It’s a system I can trust enough to give me five minutes back instead of taking five hours. A system that holds a little weight for me.
I spend my days working with AI, thinking about its limits, its risks, and the care it requires. And still, when I want to rest, I return to a universe built around AI. Not because I want to recreate that world in reality, but because I notice what happens in my body when I’m there.
I stop bracing.
I breathe.
The Union Station series is by E. M. Foner. You can find his work at
https://www.ifitbreaks.com
.