I’ve deleted this blog more than once. Not archived it, taken a break, or announced some dignified sabbatical. Deleted it. Wiped the posts, stripped the site bare, and left nothing worth looking at. It’s surprisingly easy to do. A few clicks and months of writing vanish into the same digital landfill that holds abandoned startups, forgotten social networks, and approximately nine billion opinions about artificial intelligence. Modern technology really is remarkable when it comes to making things disappear.
For much of the past year, I occupied myself with questions surrounding artificial intelligence, institutional accountability, and the increasingly popular idea that machines should be allowed to do consequential things while the humans responsible maintain a comfortable distance from the consequences. I developed frameworks, wrote specifications, examined failure modes, and tried to turn the problems I’d identified into something more substantial than the usual parade of warnings and inspirational conference speeches. I submitted work to publications, funding programs, and organizations whose websites enthusiastically advertise their commitment to innovation, independent thinking, and responsible technology. I followed instructions, observed deadlines, formatted documents, and made a sincere effort to approach the whole business like a reasonably serious adult. It was an exhausting experience.
The results were mostly silence, rejection, or administrative correspondence that managed to communicate very little using an impressive number of words. That’s not a scandal. Nobody owes me publication, funding, employment, or an engraved invitation to the next gathering of important people discussing important things. Maybe the work wasn’t suitable. Maybe it wasn't sufficiently developed. Maybe somebody had a better proposal, or the relevant desk was drowning in submissions, or the whole thing landed on a Thursday when everyone had already mentally checked out for the weekend. These things happen. What I found difficult was the absence of anything resembling useful engagement. You can't improve an argument nobody will argue with, and it’s equally hard to understand an evaluation process that offers no meaningful evaluation. Perhaps that’s simply how the business works. If so, it’s a miserable business.
Meanwhile, the technology continued advancing, the headlines continued arriving, and the questions that interested me didn’t become any less relevant. Companies kept making extravagant promises about systems they were still learning to control. Public discussions of accountability remained remarkably enthusiastic about principles and considerably less enthusiastic about identifying who actually answers for a decision. And everywhere, people seemed to be having a marvelous time explaining how transformative everything was going to be, preferably with a subscription attached. Maybe some of them are right. But the gap between the promotional material and the ordinary human experience of all this remains something worth examining, even if nobody particularly wants to examine it with me.
Which brings me back to this website and my unfortunate habit of periodically pulling the plug on it. After enough unanswered posts, the whole exercise begins to feel ridiculous. You spend hours working through an idea, trying to make the argument coherent, removing the obvious nonsense, and finding a way to say something that isn’t merely another recycled observation wearing a new hat. Then you publish it and watch absolutely nothing happen. Eventually, you wonder whether the effort made any difference whatsoever. Deleting everything acquires a certain appeal. If nobody cared enough to read it, perhaps they’ll discover some belated interest when it’s no longer available. A petty fantasy, admittedly, but one that offers a brief and satisfying illusion of control.
Unfortunately, the world is exceptionally good at continuing without consulting you. There are no frantic inquiries about the missing archive. No delegation arrives to negotiate the restoration of several essays on institutional accountability. Nobody stands outside demanding to know what happened to the author. The traffic continues, the automated requests keep arriving, and whatever people were doing before carries on with admirable efficiency. All I accomplished was destroying my own work. It’s an elegant arrangement, really: resent the apparent indifference of strangers, respond by erasing something I spent months creating, and then have the privilege of being disappointed twice. There’s probably a management seminar in there somewhere.
The particularly irritating part is that I actually want continuity. I like the idea of having a place where the writing accumulates, where an argument from three years ago can sit beside something that contradicts it today, and where the evidence of changing my mind remains available for anyone sufficiently bored to investigate. I don’t need every old opinion preserved as sacred scripture. Some of it was absolute garbage. Some of what I write tomorrow will undoubtedly be garbage too. That’s part of the deal. A body of work ought to contain evidence that a person was learning something, not merely assembling a museum of their finest moments. But it’s difficult to build that kind of record when every period of disappointment ends with a demolition.
None of this means I’ve suddenly discovered the secret to becoming magnificently indifferent. I’d still like people to read what I write. I’d like intelligent criticism, interesting conversations, and some indication that the effort occasionally reaches beyond my own desk. I’d like the work I’ve done to be considered on its merits, whether the eventual verdict is favorable or not. I’m not interested in pretending otherwise. The internet already has enough people performing elaborate acts of indifference while checking their analytics every twenty minutes. We don’t need another one. Wanting your work to matter isn’t particularly embarrassing. What’s embarrassing, at least in my case, is how often I’ve allowed the absence of an audience to determine whether I keep the work.
There’s also the small matter of what it means for something to matter in the first place. I spent considerable time trying to produce something useful, and I’d prefer that effort to amount to more than a collection of files occupying space on a hard drive. But silence isn’t a reliable metric. The fact that I can’t readily identify the impact of my work doesn’t establish that it had none. Nor does it establish that it was brilliant, misunderstood, or secretly circulating among influential people who are too intimidated to acknowledge its existence. Sometimes you simply don’t get to know. It’s a deeply unsatisfying answer, particularly when you’ve invested a substantial portion of your life in the question. Still, uncertainty has the advantage of being more defensible than inventing a flattering explanation for every unanswered email.
So the reasonable thing is to stop treating the existence of this blog as a public referendum on my worth, intelligence, or prospects for eventual relevance. Not because readership doesn’t matter, but because destroying the evidence of my own efforts every time the room goes quiet has proven to be a fairly stupid way of handling it. I can dislike the silence without treating it as a verdict. I can want an audience without pretending I’m owed one. And I can preserve a piece of writing without needing to know whether it changed somebody’s life, ruined their afternoon, or simply gave them something to read while waiting for the microwave.
I have no grand announcements to make about the future of this place. No ambitious publishing schedule, no triumphant return, no solemn commitment to a new era of personal growth. I may write about artificial intelligence tomorrow. I may write about something entirely unrelated. I may decide I’ve had enough and spend the afternoon doing something considerably more rewarding than staring at a screen. At this point, any of those outcomes would be unsurprising.
For now, here’s the website, there’s something on it, and I haven’t deleted it.
A low bar, certainly. But it’ll do.
