AI for Artistic Visualization

Reading novels is one of my great joys in life, but I often wish I could visualize a character or setting more clearly in my mind's eye.  

Now, with Synthetic Photography, I finally have a tool for this.

The Enright House from The Fountainhead

I was recently rereading one of my favorite novels, The Fountainhead, and, as always, I wished that I could form a clear image in my mind of The Enright House, a building on the East River of New York described as “a rising mass of rock crystal…straight lines and clean angles, space slashed with a knife, yet in a harmony of formation, as delicate as the work of a jeweler….”

But I can never quite picture such a thing.  I have a fuzzy, shifting, half-formed image of it in my mind's eye. 

So I asked Midjourney to create an image of an apartment house on the East River, using parts of that description my text prompt.  Here is the first grid of four options that it gave me. 


If those first four were unsatisfactory, I could have requested more and more, varying the text description itself, or the "Style," or the "Chaos" or other Midjourney parameters until I got one that satisfied me.

The Enright House with "Chaos" parameter dialed up to 80 (a bit too much!)

This is a revelation to me!  

In a few seconds, I can satisfy my lifelong yearning for tangible images of characters, places, and things that I had only a fuzzy grasp on in the past.

Here are a few more I created for this purpose:

Inspector Javert from Les Miserables

"Case" from Neuromancer

The river boat from Heart of Darkness

And for a different feel, I created the river boat as a photograph also.  Midjourney was smart enough to make it look like a photo from that time period.

The river boat from Heart of Darkness as a period photograph

I love being able to concretely visualize these things, which have always been a bit nebulous and slippery in my mind's eye.

But, you might object, how do we know that Midjourney got it "right?"  

I think that's a meaningless question.  There can be no "right," since every reader has his or her own image in mind in the first place.

And what each of us imagines is surely different from what the author imagined.

All we can do is find something that fits the text and seems true to us.

Through the process of refining an image until it satisfies me, I'm learning so much more about the possible interpretations of these characters or settings than I had previously imagined.

This is deepening my experience of reading.

It's one more way that I'm finding Midjourney to be a powerful and delightful tool.

You can get my "Synthetic Photography" course here