I’m not a fan of it either. Unfortunately, generative AI seems to be the only way I, as someone with little artistic talent, can “afford” more sophisticated graphics.

With a growing readership on this very niche website of mine, the amount of reader feedback I receive, primarily via email, but also through the community channel, has noticeably increased. This is something that brings me joy, and I’m happy to respond to everyone who reaches out, whether that’s with replies to questions, help on specific topics, or just a simple “thank you” message.

However, for the past year, I’ve been receiving an increasing number of comments about my use of generative AI imagery in some of the posts on this website. While all the comments have been in good spirit, they share one thing in common: A dislike for such graphics, along with well-intentioned suggestions to avoid future use of generative AI for cover art or inline “artwork”.

Because of the repeated feedback on this specific topic, I decided to write this post to explain myself and the situation I’m facing with this website.

Up until now

This website has been around for over half a decade now, during which I’ve dedicated considerable effort to producing original writing, photography and sometimes graphics for roughly 130 articles (not counting regular pages). Out of this content, only around 10 posts (~7%) feature imagery produced by generative AI, which I always disclose, usually in the article’s footer, and sometimes with slightly sarcastic remarks about generative AI.

However, despite my focus on original work, I’ve never received feedback explicitly appreciating the artwork featured in the remaining 93% of the posts. While the purpose of good artwork is to blend in with the writing and thus become one, it is nevertheless disheartening to see that as soon as I introduced generated images, I received immediate feedback, despite the fact that these images blended in better than any of my original amateur photography or artwork ever could.

Honestly, it’s effort

Unfortunately, many readers don’t fully realize the extensive work involved in creating both the written content, the accompanying graphics, and sometimes even videos for each post. As someone who isn’t a professional artist and who faces the challenge of finding new subjects for niche topics like “The Small Web 101” or “Installing Alpine Linux on a Bare Metal Server”, I turned to generative AI.

To give you a sense of the effort that goes into just the writing, let’s take the aforementioned article as an example: Researching, drafting the idea, expanding on it, refining rough edges, proofreading it repeatedly, and ultimately running it through a grammar and spell checker usually takes me about 30 hours for a post like that. I don’t use things like dictation and speech-to-text conversion. While new technologies might speed up the process, they would also partially take a way the joy I find in the process, and quite possible produce a result that is not as thoughtful as it might otherwise have been. Besides, I wouldn’t want to lose the ability to do these things on my own by slowly offsetting more and more tasks to computer programs. The fact that I still need to use tools for grammar and spell correction after all these years is frustrating enough.

However, with my process still being predominantly based on blood, sweat and tears, an article that requires deeper investigative work easily doubles the aforementioned number from start to finish. Keep in mind, these numbers don’t include any work on graphics!

Even shorter articles, like the “Tabs vs. Spaces” one, end up taking an absurd number of hours to complete. Manually searching through every language’s official and unofficial developer guidelines to determine whether tabs or spaces are preferred, and what indentation size should be used, takes a lot of time. I’m not using “AI” tools to automate these tasks because they simply can’t be trusted to produce accurate data. Especially with an article like this, accuracy is key, and it’s the only reason anyone would find value in it.

Tl;dr: AI slop sucks, but …

Long story short, let me be clear: I’m not a fan of generative AI either. After all, my snarky comments in the article disclaimers are there for a reason. However, after years of effort and with limited funds from donations that don’t even cover basic infrastructure costs, let alone the purchase of real artwork, I had to find a way to balance the time, effort, and, to some extent, costs that go into maintaining this website. Sadly, generative AI seems to be the only way I, as someone with little artistic talent, can afford more sophisticated graphics that support the written word and are at least somewhat pleasing (or at least okay) to the readers’ eyes.

I hope this clarifies why someone with a website like mine, that is very outspoken against many commonly considered modern technologies, sometimes employs imagery produced by generative AI. I also hope that, despite your personal opinion on generative AI, the stolen artwork I occasionally use won’t deter you from diving into the actual written content. Thank you for being here.