Back in 2020, we did a run of comics here on LowEndBox called LowEndLOLs. You can see them here.
Most were six-panel stories, and to create them I used a comic creation kit I got on some art source web site. It came with a bunch of 2-D bodies, faces, expressions, eyes, noses, clothes, etc. that you could assemble in Photoshop to make figures.
Pretty basic stuff. Along with some other compositing work, the comics were enough to get the point across but certainly weren’t enough to win any art prizes. I can do a kind of neanderthal Photoshop – enough to put words on images, tweak colors, place logos, etc. but I’ve never had any formal design training.
But now we’re in the AI age. You can create entire comics with a single prompt.
So I decided to redo one of our comics to see how it looks in AI. Here’s the original:

And here’s the prompt I used to recreate it in AI:
Create a high-res square image, comic for LowEndBox. There are 3 panels on the top row, and 3 panels on the bottom row. The name of the comic strip series is “LowEndLOLs” and the name of this episode is “Billions Per Hour”.
“LowEndLOLs” should appear as a logo in the lower left corner. Use a green theme for that.
Be sure that the two characters, Angry Man and Support Rep, are consistent visually between all panels.
Panel 1 shows Angry Man screaming into the phone. He is in full nerd rage mode and is extremely angry. He says “My VM is down! This is costing me potentially billions per hour!”
Panel 2 shows a lovely female Support Rep in a call center on the other end of the phone call. She is polite and calm and is not impressed with his rage. She asks “Sir, this is a VM you pay $1.99 per month for. Am I to believe you’re doing hundreds of billions of dollars in ecommerce per month on a 512MB VM?”
Panel 3 is back to Angry Man. He rages: “I said potentially! There’s 7 billion people on the planet. About half are males, d about half of those like sports. So if each of those 1.75 billion ordered a jersey from my store, yes, it would be billions. BILLIONS!”
Panel 4 shows the Suport Rep calmly asking: “So how much did you do in sales last month?”
Panel 5 shows Angry Man at a loss for words. It’s obvious he’s barely selling anything.
Panel 6 is split showing both characters. The Support Rep asks “How about a $1.99 SLA credit?” Angry Man says “Seems fair.”
Here’s what ChatGPT 5.6-Sol produced:

Not bad. I particularly like the StackOverflow beverage Angry Man is drinking. His facial expressions are excellent, though Support Rep’s are kind of bland. I don’t think of you all as “cheap hosting people” though I suppose we all are.
Of course, she shouldn’t be wearing a “LowEndBox” name tag – we don’t run a hosting company here. So I asked for it to be removed:

This would be better if each character faced in a different direction. I like the final panel Angry Man’s face better, and I like his shirt more.
This is the reality of the technology. Each tweak starts from a clean sheet of paper, so the ability to make small changes is limited.
So which version do you prefer? And is your decision based purely on esthetics?
Because some people will say they like the original just because it was human-made, and that’s fair. The second version certainly looks like AI, and of course I told you it was AI-generated, so some people will not like it surely for that reason.
Others will say that the second is superior because of the art. The original lacks much variety in facial expressions – indeed, support rep’s face never changes. The AI version certainly has a lot more detail.
Maybe I should just learn to draw? 😄
Leave a Reply