I’ve been trying to write more for most of my career.
Not because I enjoy writing for the sake of writing, but because explaining something forces me to understand it. If I can’t explain an idea clearly enough for someone else to follow, there’s a good chance I don’t understand it as well as I think I do.
That’s most of why I’ve kept a blog, on and off, since 2009. (I’m not proud of most of what’s in the archive. We’ll leave it there.)
The problem is that writing is expensive.
Not financially. Cognitively.
You collect the information. You organize it. You decide what matters and what doesn’t. You establish a narrative, write a draft, edit the draft, discover it wasn’t nearly as coherent as you thought, rewrite it, and eventually decide that perfection is the enemy of publication and hit the button.
For years, the number of things I wanted to write about vastly exceeded the time I was willing to spend writing them.
Recently, that equation changed.
These days I spend a lot of my time building systems made of agents. Some research. Some plan. Some write. Some review. Some challenge assumptions. Some hunt for mistakes. Some care about structure. Others care about clarity.
They aren’t replacing the work. They’re participating in it.
The interesting thing isn’t that they can write.
The interesting thing is that they can help.
Every published article still takes judgment. Someone has to decide whether an idea is worth discussing, whether a conclusion holds up, whether an explanation is accurate, and whether the finished thing reflects what they actually believe. None of that moved.
What moved is that I no longer have to personally execute every step between an idea and a finished article.
Publishing used to require a team. Researchers, editors, reviewers, copy editors, technical reviewers, publishers. Most bloggers just did all those jobs themselves, alone, badly, at midnight.
Today, that’s becoming optional.
The honest reason I’m writing more isn’t discipline. I didn’t find spare hours hiding in my calendar. The overhead between “I have an idea” and “this is publishable” dropped, and a lot of ideas that used to die in that gap are making it out now.
Full disclosure, because it would be strange not to say it: this post was built that way. Researched, shaped, argued with, tightened, and handed back to me to decide whether it was true. The deciding-whether-it’s-true part is still mine. It always will be.
So this blog is active again.
Some posts will be about agents. Most, I can assure you, will have had agentic contributions along the way. Some will be about software architecture. Some about developer tools, distributed systems, cloud platforms, GitHub, .NET, or whatever’s got my attention that week.
More than anything, this is an experiment in what knowledge work looks like when the work itself becomes collaborative, even when some of your collaborators aren’t human.
I’ve been building software for a long time.
This feels different.
And it’s worth writing about.