Yeah, I know what the world really needs right now is yet another half-baked idea about artificial intelligence. And yet, here I am talking about it.
We live in uncertain times. I don't think anyone knows truly what the world will look like in another year. I'm not even certain I know what the world will look like by the end of the winter. I've never been good at prognostication. I'm on record as saying all kinds of things about what technology will do, how it will work, and being completely wrong on it.
At fifty years old, I think I've given up on the idea of prognostication, just not within my wheelhouse. However, I am willing to speak at length about what my current reality looks like right now. I think that's the only thing that I can speak authoritatively on. And so I'm going to talk about how I've spent my last week.
In the last week, I've begun to work more and more with using artificial intelligence in order to build things. I've built:
- A mechanism for connecting my obsidian vault to my blog. And by the time I'm finished, you'll be reading the output of that.
- An MCP context server so that I can ask AI to tell me things like who's watching what on my media server, the general health and status of my media server, along with more information about how it might be faring, as well as information about the vaults that I use to make sense of my life.
- I've built multiple tools at work that I've always wanted and never quite had the chops to build for myself, which have increased my speed incredibly.
- Plans and business proposals on how my organization might better move into the future.
- A voice recognition system so that I can air out my thinking by speech rather than typing. You're also reading that now. On the one hand, it may not have the same level of rigor and quality that my typing tends to, but on the other, it opens the door to a different part of my brain than that which I normally work from, and it has significant strengths in my experience comparatively.
Everything that I'm describing here was constructed in a matter of a day, ultimately. I know how most of it works. In no world could I have written it all myself. And that's the world that I'm trying to come to terms with right now.
If you've yet to see ArtCraft, then I would suggest giving this a look. I'm using most of the tooling myself quite happily, and it's replaced tools like Krita and Inkscape that have been mainstays in my working life. What's perhaps more interesting, however, is the story on how the tool was constructed, and the speed.
If you can observe behavior, you can replicate it now. If you can describe behavior, you can have it built. If you're having a hard time describing what it is that you want, which has jokingly been how people have hung their hat as regards how they'll survive the transition that we're going under, then there's tools like Speck Kit That will help you do that as well. I've used SpecKit and it works fantastically.
It works far better than I'm comfortable with. You see, I love programmers. I work with them on a daily basis. I spent my teenage years reading their posts about engineering and quality and ideas on the early days of Usenet and Fidonet, long before the internet became a thing, and I admire the tenacity of working multiple days without sleep to hone your craft and to build a thing that works. And I share a certain sense of anxiety around the notion that having spent decades of my own life learning technology as much as I am capable of doing, of building insights that few other people seem to have feels like wasted effort in the face of the speed with which these changes are coming. I don't like the idea of taking people out of the circuit. I've written about it.
But I refused to be left behind. I will not be a whale-bone corset maker. I will continue to hone my craft, understanding that my craft has changed dramatically whether I like it or not. But you know what's funny?
I do like it. I've learned to let go and having done so, I'm now working with a level of speed and efficiency that I would never have dreamed possible for myself. And indeed, it's extremely likely that it never truly was possible for me to work this way. I've never been a brilliant programmer. I've just understood the concepts underneath it. And as it turns out, that's now all you need.