The dumbest timeline keeps on getting more and more idiotic.
Image showing a thumbs down on a job listing for Gamurs who are hiring for an AI Editor to create upwards of 250 AI generated articles per week.

Gamurs, the organization that owns several publications, just shared that they are now hiring for an AI Editor.

Gamurs owns and runs the following websites: Attack of the Fanboy, The Escapist, Dot Esports, Prima Games, Upcomer, Gampur, GameSkinny, GamerJournalist, Destructoid, Twinfinite, and Siliconera. They also laid off dozens of staff, writers and editors, across most all of their publications back in March 2023. This was a reduction of roughly 40% of their entire workforce. These layoffs happened just a few months prior to this new job listing going live.

Now, they're looking to hire an AI Editor. This job would see one person tasked with creating upwards of 250 articles per week for a pay that seems to max out at around $4.23 per article. This would be accomplished through the use of AI content creation tools, such as ChatGPT or Google's Bard or even Microsoft's Bing AI. These hundreds of weekly articles would be churned out with an emphasis on shoving in as much SEO garbage into each piece as possible.

The job listing for the AI Editor at Gamurs was shared on Twitter and almost immediately ridiculed by pretty much everyone.

This move is really beyond the pale, especially for an organization that just laid off almost 40% of their existing editors and writers. Those that weren't laid off from any of the Gamurs owned publications should either be protesting this or actively seeking new jobs.

Even this new job, AI Editor, pays absolutely nothing. It's insulting to those laid off, to those still working for Gamurs, and for anybody seriously considering this position. The upper end of the pay scale says the person will make $55,000 per year. The lower end being $40,000 per year. Imagine debasing yourself to the tune of 200 to 250 articles per week for that little amount of money. And it's not exactly super easy to crank out what is roughly 5 to 6 articles per hour even with AI.

Remember back on April Fools this year? Everything that was posted that day for Wrestlin' Nation's resurrection was created purely using AI. That includes the entirety of each article's text content, every image within the articles, the avatar images for all of the fake authors, and even the banner image (though I did use Photoshop to combine different AI generated images together for that one). I posted something like just seven articles and they were not easy to create if you had any sort of specific goal in mind.

A significant bulk of the text generation for those articles was done in this very lengthy ChatGPT thread. You can see how several similar prompts can have wildly different results. You can see how sometimes the AI would refuse to work or just time out. You can see how it would sometimes not even follow the prompts even after several attempts.

The image creation was also done through AI, as noted. It was accomplished through a locally hosted release of Stable Diffusion. It required a couple hundred GBs of different models (data sets that Stable Diffusion pulls from to create the final output) downloaded along with quite a bit of time for each image rendered on my GPU, with more time needed for the images I created as I went for a "larger" resolution of 1000x563. Each image required several iterations on a theme, very precise wording, very specific details entered in on what you did and did not want included in the results. Even then, the results were often not what you wanted. This article with an AI generated "photo" of Andre the Giant? I think I tried half a dozen different models, and something like 40-50 different renders before I found one that was close to what I thought was usable.

This "photo" of John Cena was also a very time consuming project, though it admittedly was a little less time consuming than Andre was. Yet you still needed to use inclusive words such as "john cena, standing, on a movie set, wearing a soft cotton shirt, the shirt is blue, realistic proportions, realistic eyes, realistic facial features, realistic lighting, dslr, photorealistic, photo realistic, photo, photograph, dark background, soft gaussian blur, realistic eyes, realistic pupils, canon eos 90d, nikon d850, etc." You get the idea. And it also had to include things that you did not want such as: "cartoon, fake, exaggerated features, detailed background, painting, paint, fantasy, disfigured face, disfigured features, etc." You also sometimes had to add "weights" to some of the prompts to ensure that the AI generation either placed more emphasis on those features or less.

All of the images, even the obviously fake ones, took way more time to create than I care to admit. I'm talking about maybe a couple of dozen hours over the course of a week or two. Then you factor in the time for the text creation and everything else. It adds up.

It was a nightmare. I undoubtedly put more effort and time into those fake articles than if I had simply wrote them all from scratch in the first place. What I'm getting at here is that AI generation isn't just some super easy thing if you are going for quality output.

I suppose when your goal is just to churn out upwards of 250 articles per week, you aren't concerned with quality in the slightest. You just want to oversaturate search results everywhere and the quality of your output be damned.

I used AI tools simply for a joke, for resurrecting an April Fools blast from the past that we had before here with CS-Nation. There isn't a single time where I would ever seriously consider using AI generation for actual content output. Sure, you can have way more content going up at TGN here if I did that, but it would all be trash. I mean, it would be more trash than my usual garbage. Every article would just be this super formulaic nonsense that has no heart or soul. Each article would be like all of those other ones you find online where there are like 3-5 paragraphs of the most generic information about a topic before it finally gets to the point of why you stumbled on the article in the first place.

Fuck that. And fuck literally any company that seriously thinks about going this AI generated content route, especially after laying off dozens of employees not even three months prior.