Transparency: How I Use AI in the Age of "AI Slop"

This is the English version. Read the original: Deutsche Version

Since we have collectively decided to clutter our internet with AI Slop, I want to briefly outline how I have used and will continue to use Artificial Intelligence in the future.

I write my texts myself; I neither provide an AI with an outline nor do I have articles generated [1]. Also, images, unless otherwise labeled, will be accordingly real.

However, since I am not a Luddite and do not condemn AI in general, I still use it for the following use cases:

  • Spelling correction, punctuation, and grammar.
  • Translations.

Why? For one, because I have a complex relationship with the german "dass/das" (I guess in english it is "that/which"?). I often rewrite sentences but don't always completely remove fragments of previous versions, and it simply makes me feel better. I am relatively confident that current AI can handle this reliably.

And translations: simply to save time. I can read, write, and speak English fluently, but it still takes time. To keep my motivation high, I want to put as few obstacles in my own way as possible, which is why the texts are translated either by an AI or DeepL.

Should I use an LLM for anything beyond what has been mentioned, I will label it accordingly. And since I couldn't pass up the opportunity to demonstrate this right away: The following image is definitely AI-generated.

THIS IMAGE IS AI GENERATED!!!!

Footnotes

  1. Note: I did do it once, back when it was new. The quality was poor and it felt soulless. I won't say which text it was, though—maybe you can find out.