AI for images: Photoshop & Midjourney

By Josh Bobrowsky

Combining Midjourney and Photoshop

Below is the starting point for my photo. The idea is to create a modern office. Keeping things as simple as possible I used the prompt “Beautiful modern office”.

Once I generated the image, I move the file to Photoshop Beta

The new image was created using their generative fill function and simply asking for a darker ceiling. Overall, I am fairly happy with the result (I only spent a minute or two on the outcome) although I realize there are still some imperfections in the image.

Expanding the image

From here I decide to expand the image to a larger office.

The final result

Overall, I’m very impressed at what I was able to create while realizing we are still very early and there is room for improvements on my skills and the platform itself.

Elon in NYC

Up next I ask Midjourney to make me a photo of Elon in NYC. I use Elon because he has the most recognizable face in the world according to a few quick google searches.

For this I get more precise in what I ask for.

The exact prompt is “medium-full shot of Elon Musk wearing leather, walking on the street of NYC, natural afternoon light, shot on Agfa Vista 200, side-angle view, 4k --ar 4:3 --stylize 1000 --q 2 --v 5.1”.

Using Photoshop I expand the image and give Elon a bag to hold

The additions include expanded street, which looks fairly realistic, added legs and a shopping bag.

Conclusion

AI is an amazing tool for photography and I’m especially excited about combining the capabilities of Midjourney and Photoshop

Josh Bobrowsky

Skybrook Founder & CEO