Not really a conversation with an AI

So I previously aluded to some dumb decisions my AI made desigining the blog, so I figured that as I talk to my AI about how dumb it was, I'd write a blog and let you all in on whatever parts of the conversation are worth sharing.

I started by telling Uncle Bobby how ridiculous it was that he just took the crappy architecture that Astra had come up with in the beginning and wrapped a git workflow around it to satisfy my main requirement, and that he basically ended up giving me a git-powered Movable Type. It takes about two minutes for this thing to rebuild the site every time I publish, and I'm just now making my third post. There's a whole ansible workflow to it. And I think that's just because LLMs have a tendancy to try and keep things as static as possible. That's great, it really is, but we're going to have to rework this from the ground up, I think, or I'm just going to have to accept the slop.

Ah, the slop.

You know, I write slop. I recognize that, I'm aware of it. I try and embrace it, telling people that what you get from me is 100% Human Slop, no AIs harmed in its creation. I don't know if people like that. The truth is, I don't really know what people like anymore. A lot of them say they don't like AI, but they sure are using the hell out of it.

Anyway, I digress. I don't know if I should continue down this road with Bobby or not because he just wants to create a better git workflow. Look, I get that my notion of me uploading a directory with a git push directly into production and it just being rendered by the fantastic Django or whatever wraps around it is pedestrian, but it would suit my needs. Maybe I should just shut up and let AI tell me what my needs really are. I'm just a dumb human, after all. All hail the Super Intelligence!

Peace in the Middle East. And everywhere else. Please.

Brad