A Geek's Life

Cafélog, who is on my blogroll, responds to a post at Vrypan.net, AI-generated blog posts images are not cool anymore . It is a detailed response which delves into their use of AI and the process involved in generating the images on the blog. As someone who also uses AI to modify pictures for use on this blog, as I mention here, it is difficult to not feel similarly affronted by that view.

The central point of Vrypan’s original post is that “everyone does it”, with the result being a visual fit of a “2,000 Power-point using animated transitions and WordArt” with a few exceptions. Whilst that may be true, I do not necessarily see that as a problem. Power-point persisted (and still persists) in the office environment for a reason, which to my mind lies in the confluence of availability, a low barrier to entry and some may say lock-in – all features of LLMs today. That lock-in is bad is non-controversial. Availability and a low barrier to entry though democratise (commoditise?) the means of production, blog post images in this case.

For me, clarity around how one uses AI is key. Folks who then come across the blog, a specific blog post or perhaps interact through an OpenGraph snippet then have the choice of engaging with the content or otherwise. Generating a /ai page is now on my agenda for the next month.

Written as part of the JulyReply series