Lensa is just the latest photo editing tool to drive the internet into a frenzy. It appeals to the narcissism within all human beings as it creates "avatars" based on pictures users upload to its servers. But it's far from the only AI that's being used to render images.
Just today I discovered that the editing software I use for images I include in my articles had rolled out a new "text to image" feature that promises to create images based on prompts you give it. I used it for a previous article I wrote about tattoos just to see how it worked. It turned out nicely so my next move was to find its limit. I gave it a prompt to see if it could replicate my hometown, Tri-Cities. I was surprised to find that the AI has a good idea of what this area looks like. For each prompt, the AI will create four images.
My first prompt was for Kennewick. I think the AI did pretty well.
You can see the AI knows there's a river here and at least one bridge that goes over it. The next prompt I gave the AI was for Pasco. You be the judge.
I think it looks pretty similar to Pasco. Note that the AI didn't revert to any of the images it had given me for Kennewick. So, what about Richland?
The Richland images look somewhat similar to Kennewick but still get the general essence, even if it's missing a certain George Washington Way. I then asked the AI to render Benton City.
Holy cow, the AI nailed Benton City! But what about West Richland?
I thought West Richland would be tough for the AI since Richland was in the name but I was surprised by how well the AI replicated it. That's when I decided to go all in and asked the AI to render the Hanford site.
I think the United States government should have a talk with this AI since it knows too much. All jokes aside, I thought this was a fun exercise. Just remember to be nice to the robots of this world since they will be ruling over us one day.