It’s not hyperbole to point out that many if not most photographers hate A.I.  When the term “A.I.” is brought up to most photographers, they instantly think of Generative A.I being used by potential clients that use it to produce images instead of hiring a photographer.  That fear isn’t entirely unfounded.  As photographer Shane Balkowitsch recently wrote in PetaPixel, “Real photographers are being pushed aside and A.I. image generation is now a major source of digital image production worldwide.”

Sure… some photographers are losing out on business due to A.I.  Naturally, that’s going to create some resentment among photographers.  But photographers need to be careful not to throw the baby out with the bath water.  A.I. can be used to make your camera-produced photography genuinely better. More importantly, it can help make you a better photographer.  To illustrate this point, I’m going to demonstrate how I used A.I. to shoot a featured image I needed for an article that I did here recently on ThruMyLens.

The article in question was a comparison between the DJI Osmo Pocket 3 and the Insta360 Luna Ultra.  I needed an establishing shot – a “featured image” in WordPress speak.  I could have shot both devices against a white background and called it a day.  But I was already using several plain white background “product shots” in the article.  Using one as the featured image wouldn’t be very interesting.  I wanted to do something more interesting.  But what?

Personal confession:  I’m not naturally great at coming up with creative concepts from scratch.  And often, even when I do think of a good idea for a photo, I don’t always come up with the most creatively exciting way to shoot it.  So I decided to try and bounce some ideas off ChatGPT.  I happened to use ChatGPT for this exercise, but there’s nothing particularly platform-specific about the process. Gemini, Claude, or any number of other conversational AI tools could be used in much the same way.

FILLING IN THE BLANK SHEET OF PAPER

I started off by describing a somewhat vague/general explanation of my problem to ChatGPT.  “I need to create a creative, interesting featured image for my WordPress article on the DJI Pocket 3 vs. the Insta360 Luna Ultra.  The two devices should be the focal point of the image.”   I say “somewhat vague” because I was specific about certain things – the image was intended to be a WordPress featured image, it needed to be visually interesting, and I specified that the article subjects should be the focus of the image.  Beyond that, I wanted to see what ChatGPT could come up with as a concept.  Here’s the first image ChatGPT generated:

ChatGPT knows I live in Ohio so it chose to depict the two cameras in the foreground with the Cincinnati skyline in the background.  This was an interesting thought – get the products out of the studio and do something outside.  But I quickly dismissed the idea – too much time and effort.  I wanted something I could shoot in my studio.

I then started thinking about what props I could incorporate into a photo of the Pocket 3 and Luna Ultra.  Maybe I could throw my Fuji X100VI in the background?  Maybe some of my vintage Canon FD lenses?  I also like throwing in a watch into my photos whenever I can – I’m a watch guy.  So I told ChatGPT, “I think we need something that looks like it could’ve been taken in my studio. In the background, it could have a Fujifilm X100VI, a modern Rolex submariner, and maybe some FD lenses.”  Here’s what it came up with:

There were all sorts of “A.I. slop” problems with the image ChatGPT created – but a concept started to take shape.  These two gimbal cameras are often used when traveling?  What if I leaned into that and add elements to the photo that were travel themed?  So I got my passport out of the safe, as well as a small globe I’ve used in other product shoots, and started placing all the “props” on my staging area.  Behind them I used a white vinyl background, knowing I would replace it in post.  I then grabbed my Westcott FJ400 II and my Westcott 36” Octa-M and took a test shot using my Canon R5 Mark II and my Voigtlander RF 40mm f1.2:

I uploaded the test shot to ChatGPT, which then gave me some great suggestions about making subtle changes to the placement of the objects on the stage – things hadn’t noticed on my own but made a big difference.    I took a few more test shots and played with different aperture settings.  I then asked ChatGPT, “Give me a mock up of how you think the composition should look.”  Here’s the image it generated:

I noted that ChatGPT placed a pen and a leather notebook in the bottom left hand corner of the frame with a tiny bit of what looks like a green houseplant sticking out.  I didn’t have any houseplants handy, but pen and notebook inspired me to grab one of my Rolex pens and a leather Rolex wallet out of a display case I have in my office that showcases my Rolex collectibles.  I took another test shot at f5.6:

I had to use my finger to hold down the edge of the passport, which I knew I could remove in post.

At this point, I was ready to take the final photo – I turned on both gimbal cameras, pointing them both directly at me:

From here, I just needed to do some editing in Lightroom and Photoshop – I changed the background (adding a stream of window light), cleaned up the lines on the wood surface, and I blacked out the screens on the gimbal cameras.  Here’s the final product:

Seeing the final photo gave me a huge feeling of unlocking a creative secret.  The image was largely done “in camera,” but I used A.I. as a collaborative partner to create something I would not have been able to envision on my own.  I felt as though I had discovered how to fill in a creative gap that’s been holding me back for years.  It was an amazing feeling of both inspiration and accomplishment!

It’s important to point out what A.I. didn’t do.  It didn’t take this photograph.  I chose the camera and lens. I arranged the objects.  I positioned and powered the strobe.  I chose the aperture.  I composed the image and pressed the shutter.  A.I. gave me ideas, pointed out things I hadn’t considered, and gave me something to react to.  In other words, it didn’t replace the photographer—it helped the photographer.

Hopefully my sharing of this creative collaboration with A.I. has given you some ideas as to how you too can use A.I. in your photography.  I would encourage you to, whenever possible, upload photo concepts and ideas to A.I. and get feedback – let A.I. identify your “blind spots” – let it make suggestions and point out possibilities you wouldn’t otherwise see on your own.  A.I. is a tool – let it help you get past the blank sheet of paper and unlock your creative potential.