I just created 5 images with Google Gemini — and it left me both impressed and annoyed

[admin]

This hands-on experiment takes a look at the image generation quality of Google Gemini's Imagen 3.

I just created 5 images with Google Gemini — and it left me both impressed and annoyed : Read more

 

[Burriteo]

This article is just... bad.

Prompt: “Create an image of a cozy living room.” I decided to be a little vaguer with this one to see what Gemini would create. The result was hilariously bad.

When I think of “cozy” I want a big comfy couch, big windows, soft lighting and a warm blanket.

If I ask 10 different people what is their idea of a cozy living room, I will most likely have 10 different descriptions. Would I tell any of them that their idea is "hilariously bad" if it doesn't match with MY idea of what is cozy? Coziness is subjective anyway, so why would the writer expect an AI image generation to have the same idea as the writer? It can't read minds unless one describes exactly what their idea of coziness is like.

However, it’s clear that it needs extra prompting to truly nail down the final visual. It struggles with more nuanced requests and can’t seem to creating unique, artistic interpretations.

This is how most of AI generation works and if one wants better results, write a better prompt with clear specifics and details. This feels like criticising for the sake for criticising. Is the writer even knowledgeable, experienced in this subject matter? Or better yet, has the writer actually used one before writing this article?

 

[Zantie]

Terrible terrible article

 

[Tom25!]

This hands-on experiment takes a look at the image generation quality of Google Gemini's Imagen 3.

I just created 5 images with Google Gemini — and it left me both impressed and annoyed : Read more

This article is as terrible as the writer's ability to use AI. The article author's title is "AI Writer" but it's relatively clear she's never written anything for AI in the past. These are the worst prompts I have ever seen.