TL;DR (that's "Too Long; Didn't Read" — and yes, we're aware that explaining the acronym makes this longer, which is probably the most honest thing on the page):
We use AI. We're not hiding it. We run most of it on solar power at our house instead of in a data center, we never use it to lie to you, and nobody here posts raw AI output without a human tearing it apart first. That's the whole policy. Everything below is just us showing our work.
The entire AI industry right now is built on one idea: more. More data centers, more power, more scale, bigger models. I think that’s mostly wrong, and I’ve done the work to back it up.
I’ve written about what I call constrained AI — the idea that you get more reliable, safer, more accountable results not by chasing a more powerful model, but by designing the workflow and the limits around it properly. I’ve built processes that run perfectly well on the AI models that already exist, and that actually perform better in organizational and business settings — more safety, more security, more accountability — than what the big companies are selling. (Paper, supporting material, and live demonstrations available on request. I’m not asking you to take my word for it.)
We don’t need more powerful AI. We need better and more ethical use of the AI we already have. The current models are good enough. People are just using them badly.
You know the stuff. The obviously-generated image with seven fingers. The email that reads like a press release wrote it. The website copy that says nothing in 400 words. That’s not what we do, and I’d ask you to hold me to it.
Everything we touch with AI is monitored, controlled, written, revised, revised again, and gone over by actual humans before it goes up. My wife works with these tools and knows them cold, so nothing sloppy gets past her. And if we’re ever tempted to cut a corner, we’ve got a built-in ethics board: our two kids, who despise AI with the white-hot intensity only kids can manage and who will absolutely let us hear about it. Living with your harshest critics keeps you honest.
Here’s the part people don’t believe: a lot of the time, AI doesn’t even save us time. When somebody sends a long question through this website, I write the whole answer out myself, the way I always have. Then I hand it to a custom AI tool and ask it to rewrite it the way a person will actually read it — and it cuts my rambling three-page reply down to half a page and ten bullet points. That doesn’t save me a single minute. What it does is make sure the thing I’m trying to say actually gets read and understood. That’s the point. Not speed. Clarity.
We’re not here because AI is amazing or because it’s “the future.” We use it because it’s a tool that lets us do things we genuinely could not otherwise do, because of, well… life.
We have two kids. We have four cats (not by choice; a hawk dropped one in our yard, and here we are). We have real health issues in this house that eat the hours we’d rather spend on other things. Add a high-stress, high-impact day job, a community we run to support people of every type and belief, the occasional desperate attempt at a date night and a movie, and a house and yard that are perpetually one project behind. (If anyone knows someone who does plaster work cheap, please, I’m begging you.)
We’re not amateurs at this stuff. My degree is in UI/UX with a specialization in graphic design. Both of us can do art, editing, video, and audio at a professional level. What we can’t do is everything. We can’t afford to hire people to do it for us, and honestly we can barely afford the AI tools we do use on a monthly basis, never mind a subscription to something like Adobe Stock. (Have you seen those prices? And half of what’s on the stock sites now is AI-generated anyway.)
Here’s the line we will not cross: we’re good enough with these tools that we could generate images and rework them so they don’t look or read as AI, visually or in the code underneath. We never will. Even when it would save us time. Because that’s dishonest, and not making things worse by lying to you is the entire reason I’m running.
This is the part I’m proudest of, and it’s where the data-center argument stops being talk.
We’ve been shifting our AI work toward sustainability, with an environmental focus. and these days the majority of it runs locally, on our own computers, right here in my office. Most of my AI work doesn’t touch a hosted data center at all. Once they’re set up and trained properly, my machines at home generate images, handle editing, and produce text about as well as the cloud models did a year ago. And because my whole approach was built starting from those local constraints, it’s structurally sound from the ground up.
I still use top-tier cloud AI, but not to do the work. I use it as a development and management layer that tells the AI running on my own computer how to do what I need and controls those processes, which actually makes my local output better than it used to be. (Again: papers, documentation, examples, and presentations available on request.)
And to power the extra load, we’ve moved to solar and scoped our setup to handle it. So when I say we don’t need massive data centers, I mean it literally, I took the majority of my AI work off them.
(And yeah it was a bit of money to spend but not as much as you would think)/
We’re trying. And what we’ve learned is that where AI already is, today, is good enough for most work people use it for. People just aren’t using it effectively, efficiently, or with any focus. The workflows are poorly designed. The hype is, like most hype, a pile of half-truths and lies dressed up as inevitability.
We don’t need the massive data centers. We don’t need to push people out of their jobs. We don’t need to blindly follow whatever AI spits out, or whatever the people selling it tell us we’re supposed to do with it.
But we also can’t pretend it’s going away. Pandora’s box is open. And here’s the part that should worry you: the only way to truly shut it back down now is to take it away from every regular person out here, and leave it solely in the hands of corporations and the government.
That’s not safety. That’s a handful of powerful institutions keeping the most powerful tool of our era for themselves while telling the rest of us it’s too dangerous for our hands.
That is a very long way from real freedom. And freedom is the thing I’d actually like to defend.