Were AI Blog Post Images Ever Cool?

I read an interesting blog post by Panagiotis Vryonis titled AI-generated blog post images are not cool any more. My first thought was “they were never cool.” But I read on to see his perspective. While I and many others serve as proof that asserting that “everyone does it” is flippant, he does offer a practical point (from a non-AI skeptic perspective) that enough people use AI images to illustrate blog posts that it has the effect of “look[ing] like a 2000 PowerPoint using animated transitions and WordArt.” Mr. Vryonis offers a sensible AI-powered alternative that is less ocularly offensive than many of the AI images and would not raise that same hard-to-suppress tendency to question whether what follows is genuinely authored by a human being (commenter cmc highlighted that issue well). But I like to be solutions-oriented, so I offer my own take on the issue.

Back before I launched The New Leaf Journal in 2020, I thought it would be fun to use my own photographs and public domain sources such as Project Gutenberg books for photographs and images on my site. As people who frequent my posts know, I have stuck to that approach for more than six years, although some of my articles are illustrated by screen captures instead.

My photograph of two pigeons in the street in Brooklyn Heights on April 10, 2026. Straight from the camera lens of my LineageOS-powered Google Pixel 3a XL (subsequently cropped and compressed for publication).

While some people may have reasons for not wanting to publish local photographs, I appreciate seeing writers put some work and a human touch into their illustrations. Because Mr. Vryonis also references using images with creative commons licenses, I will note I have done that in a small number of posts, using images from Wikimedia, Flickr, Pixabay, and Pexels with the appropriate licenses, but in general I try to stick to my original approach. I would personally recommend against using AI-generated images, both because the sort of flattening effect Mr. Vryonis describes and their tendency to give off an inhuman vibe about the post as a whole (deserved or not), but I do read some blogs that use them (see Marius’s contrary On Generative AI Imagery take on his interesting blog) and would not go close to so far as to say they raise the same questions as writing bearing obviously inhuman AI markers.

You can reply to this article from your own site by sending a Webmention.