Stop With The Slop

Adam BieganskiAdam Bieganski
2 min read

Recently my colleague Toli included this disclaimer in an RFC:

🤖 I’ve written this RFC with the help of ChatGPT for wording (hopefully easier for you to read!), but I’ve read it all, and they are intentional design decisions. If you do see anything that’s off, add a comment.

I felt relief and appreciation: “How thoughtful of him to put me at ease like that. Maybe we should all do that?”

See, the problem isn’t that I wasn’t expecting a thoughtful, relevant RFC actually worth reviewing — especially coming from this author.

But I’ve felt recently that the amount of AI-generated slop everywhere is so high that I constantly have a background process running in my brain that is trying to detect it. And the more often the process raises an alert when I’m one-third into some text (“Hey, wait a minute, this is AI-generated slop, I’ve just wasted so much time and attention…”), the more difficult it is to suspend that background process — which is very heavy on my resources.

Incidentally, today I stumbled upon this post:

https://www.seangoedecke.com/dont-feed-me-slop/

which made me realize that I’m not alone.

So it seems like the main thing may not be the revealing that AI was used: “I’ve written this RFC with the help of ChatGPT”, but the second part:“but I’ve read it all, and they are intentional design decisions”.

The post also let me give a name to that feeling when the scales tip: “the unpleasant uncanny-valley sensation when you realise you’re not talking to a human”.

P.S. I didn’t use AI when writing this post, but I deliberately used em dashes for punctuation above, to trigger your AI-detection background process. The cover image was not AI-generated: https://www.flickr.com/photos/rexroof/3267476203

1
Subscribe to my newsletter

Read articles from Adam Bieganski directly inside your inbox. Subscribe to the newsletter, and don't miss out.

Written by

Adam Bieganski
Adam Bieganski