Here’s my attempt:

Prompt: Create a picture of a first person point of view: standing on a stainless steel walkway. Directly below the walkway are large cubic water tanks. These tanks are made of stainless steel and filled with murky water. Within the water, there are subtle, barely discernible shapes of ominous creatures, adding an element of mystery and foreboding. The walkway is a 3x3 metal grid, perfectly centrered and suspended at a mid-level height in a large, sterile white cubic room. This room is brightly illuminated by harsh fluorescent lights from above.

= = = =

I noticed that your prompt is a bit messy in regard to the features of the room you described: I found that it might be easier for the AI if you describe the things you want to see very separately and one after the other in a logical order, like you would paint them.

You wrote “stainless steel walkway, perfectly centered and suspended at a mid-level height in a large, sterile white room.” Then you describe the room further, then you describe the walkway again, then the water tanks, and then again the room’s walls.

I think it’s better to focus on the walkway above the tanks first since they are the central objects, and then place them in the surrounding, the room. So focus points first, then the background. I’m not sure if this is understandable, but I hope it can help to clear up something about how the AI might struggle to understand your description.

My prompt is not very clear in this sense either, but maybe trying further in this direction helps.

I didn’t manage to get it to create a cubic room either :-)