In these times of AI-generated content it is far too easy to run into slop. I would claim that in some parts of the social internet AI-generated low-quality content is prevalent. When deciding on an approach to posting here, I had some reservations.
It has become far too easy to get an AI assistant to explain how things work. It is even a very enjoyable activity, to learn by grilling the chatbot to go as deep as desired into the technicalities. At the same hand, it would be far too easy for me to prompt whatever service I am using at the time with something like:
Write a blog post about the how to set up a playground of [insert a selection of services here] that interact amongst themselves. Use [insert list of technologies here]. Here is a list of similar posts that I have written in the past for you to imitate my writing style.
Therefore, I see very little value in doing either of the two.
- Anybody can generate something similar, so the bar is lower
- Very few people might be interested in reading. This is because there are better learning options out there plus a lot more cheaply generated content competing for reader’s attention.
Finally, we no longer google & browse to find this type of content; we ask the AI straightaway.
So, is there room for making technical posts in this post-LLM slop world?
I guess that my answer for now is Yes. By writing content that is short and straight to the point, by hand, and without letting an AI assistant interact with it. This is perfectly summarized by Jonathan Stark’s 100% typo guarantee disclaimer: the content is made by a human, without an editorial process, and therefore it is guaranteed to have some horrible typo which bears witness to its authenticity.
Header photo by Álvaro Serrano on Unsplash.