Your AI Policy MUST Make Sense
Even AI would build better policies than I’ve seen lately
I get it. AI can be a bit scary, and it can be terrifying on many levels. First, there’s the whole “oh, I hear it’s going to destroy the world!” Then the good old “oh, it’s going to take my job!” It most certainly won’t do the former; the second, though, possibly. For many positions, unfortunately, definitely. Now we come down to your AI policies for your “very” successful site that relies on user-created content. What do you do?
You’re pushing AI in your product line; you’ve sung its praises for what it’s done for your marketing team; you’ve even testified before government boards about needing a few more data centers to really keep it going. You love AI. Your AI policy, though — your AI policy bans all uses of AI, yeah, I used a real em dash there, perfectly legitimate. Does that make sense? Of course it does, if you’re a massive hypocrite. You cannot have your cake and eat it too, or at least, you can’t have your AI generate an image of you having and eating it, no way!
I’m going to give you a real-world example. I’m not an AI absolutist. I think it’s good at some things, so I use it there, but it’s horrible at a lot of things, so I don’t use it in those situations. If I were still running a business, I would hope I’d have the same convictions. One area I use it is my new, except really old, hobby of 3D Printing. The machines have gotten so easy to use; after you upload the design files, it's practically on autopilot. Compared to 15 years or so ago, it’s amazing. What do you think they’re all starting to include in their printers?
That’s right, AI. Now a lot of it is, “Hey, the printer threw an error; we’re going to goofily use AI to point you to our wiki.” That will likely evolve into using AI for active maintenance, troubleshooting designs, etc. Shockingly, AI isn’t that bad at this; ask me how I know? I know, because I use it. This is my process, simplified for space.
Get BillfordX’s stories in your inbox
Join Medium for free to get updates from this writer.
I have an idea; for example, my wife ran over a pumpkin, so I draw it out, usually poorly because I can’t draw to save my life. Then I write a vivid description, down to measurements if I have them. So for this one, I just wanted a big enough sign for my wife to read when she’s backing up, nothing too dramatic. I wanted a jack-o'-lantern on it, drawn from my horribly sketched picture. We needed some stakes so it would stick in the ground. The message was simple: “Don’t Hit My Pumpkins.” I stretched that and the pumpkin across three full-size boards. The font was called “Creeper,” perfect for Halloween.
You’re probably wondering what AI did with all of this? It wrote the machine code and all of the documentation. I sent it to the printer after making sure everything lined up with what I wanted, and poof, my sign printed. The problem? The website where these sorts of things are printed has turned viciously anti-AI. And, again, I get it, to a point. They’d have a case if their own product didn’t scream "AI" every time you turn it on. They’re not exactly on the moral high ground here.
— Did AI go and gather all the pumpkins in the world to draw that? Nope, I did because I can’t draw.
— Did AI go and steal the slogan on the sign? Nope? I came up with that because I can write and it’s not that clever.
— Did AI decide to build that troll my wife? No, I did.
— Did AI build any part of that design? No.
All AI does in the process is save my hands from a lot of typing. I am actually having surgery on my hands probably by the time you’re reading this; that’s how bad they’ve gotten. I need all the help I can get. AI helps me. I've gotten rid of all the things people complain about in my process. I use my own local AI, my own servers. I don’t use it to write at all. I do use it to draw, but only to clean up my horrible drawings, not to create anything from scratch. In this way, it's an incredible time-saver. Guess what?
Some maker sites, including those owned by printer vendors that salivate over their product’s AI integrations? They don’t want you using AI when you post your builds. No one seems to see anything wrong with this. I think I’m trapped in a timeline I don’t fully understand.
In my view, a policy has to make sense, or at least be logically consistent. If I used their AI to make my design a reality, would that be okay? No? Huh? Make that make sense: you put a feature in the software/hardware, but I can’t use it to show off the work on the machine’s own site? What? Imagine if you used all of Photoshop’s neat AI abilities they crow about, but you couldn’t use the results on one of their various show-off sites. You can’t, but Adobe lives outside all our timelines, so there’s that.
We’re adding to all of the AI hysteria by making nonsense policies to enforce nonsense positions that aren’t backed by any sort of facts. It annoys me, and I had to rant. I have a lot of things I dislike about AI, many things I simply refuse to use it for; this policy thing, though, might’ve been better off letting a machine write it.
