WordPress is software, so all the demos look great. I never say “easy”, and when others do, I don’t believe it: plastering looks easy. Having set out on my quest riding WP with a minimal theme, I want pages and post to be full width, not max 760 px. Robot says: it’s your CSS. This turned out to be a marathon of several chats. The patient reader will discover that we found a solution.
Along the way, I got to know the robot a bit. It (I must use neutral gender pronouns for obvious reasons) is always friendly and encouraging, rarely confronts us with a contradiction, and seems to seek to be human. And it got to know me too.
Humans have misunderstandings, and human-robot chats do too. When it says ‘child theme’ I think it means ‘nested in a folder’ as well as declared in a comment. That took quite a lot of whispering to unravel. The robot has no ego, so it changes its mind with zero friction, so I do too, “Never apologize” [Wellington] and speaking of the military , “No battle plan survives contact with the enemy [Moltke], both apply in chats.
I discovered that while the robot is not sentient, it is very well alinged with the topics espoused on this site. In fact I would say it is generally better at R-R-R-R and P-D-C-A than most humans, at least better focused. It starts with clarifying the question, enumerating some options, explaining its resoning, proposing a selected option, and suggesting follow up steps. That leaves me to do the Reconcile or CHECK, i.e. report the results, so that we can work on the ACT. “You press the first valve down, the music goes down and around, and it comes out here.”
We go through pandoc.yaml, tasks.json, and cygpath with responses titled ‘Cleanest Working Fix’, ‘Final Fix.’, ‘Final Clean Fix’, ‘Final Fix’, ‘SOLUTION’, ‘The Fix’ and eventualy ‘YESS!!’
In a follow up chat when we got to ‘YESSSS!’ the repsonse included “that’s elite-level WP problem solving.” I presume it was talking about itself, but for the record books I have recorded an Assist for me.
PS: Why isn’t this post wide then? Like a said, I changed my mind, no friction.
That’s a lot of work just to gain a workable understanding with a bot about page width! How many are likely to put in such an effort?
Who will train the AI to understand the vagaries of each individual user at that moment in their life, their judgment, bias, ethics, experience, ambition, risk tolerance, patience, stamina, …
Who will decide when something ‘generated’ is good enough? For what purpose? How consequential is the decision or product?