In my last tip, I wrote about checking for accurate digits and limbs in AI illustrations.
Today, I would like to simply do a comparison of the images one might get from four of the top AI platforms: ChatGPT, Gemini, Grok, and Claude.
Here is the prompt I used for each platform: Create a cartoon illustration of a young husband and wife in a contemporary furniture store. They are talking to a female sales rep about 30 and attractive. She is showing them an easy chair. The husband is trying it out and the foot lift is up. Colors should be vibrant and warm. All look happy.
Here are the results.
The above ChatGPT image is crisp and followed the prompt instructions precisely. It’s a little busy, but I like it. Last year, the images from ChatGPT were fuzzy. I guess they’ve improved.
The Gemini image above also followed the instructions precisely. It’s not as crisp as the prior one. This is interesting because I used to prefer Gemini because of its “crisp” images, but lately they’ve had a softer look with busier backgrounds.
Grok did well, although it didn’t follow the instructions as the man is sitting on the couch rather than the easy chair. If I were to go on with it, I might lighten it up a bit.
Oh, dear.
I prefer cartoon style illustrations as they are obviously fake. Lifelike AI illustrations tend to look too perfect, plastic, or like creatures from a wax museum.
Easy Web Tip 385: If using AI to create images, play around with different platforms to see which suits you best.
