Over the span of the last couple of weeks, I built out this site and I relied heavily on AI.
All of the images except for my self-portrait were created by Midjourney. I have my own server set up so I can create and track images (check out my LinkedIn post to see how I did this). It took me a little bit to figure out how to prompt the AI so it would output images I was happy with and they all had a similar feel to them.
As we know in branding, consistency is key. You’ve most likely been a designer long enough to understand the importance of consistency. I actually manipulated my self-portrait a little so it fit with the other images on the site.
There were a couple of pages where I needed wide shots, and the images I had selected didn’t quite work. That’s when I brought the image into Photoshop beta and tried out the generative fill capabilities. It blew my mind how easy this is.
You’ll also notice that the original coffee mug had two handles on it. A simple lasso and a prompt telling Photoshop to get rid of the handle and *poof* I had the image I wanted.
Most of the copy on the site was created in ChatGPT. I’m writing this blog post manually, but a lot of the other pages were generated based on my prompts. I’m still getting better at building tone into the writing, so I’ll probably go back and update pages as my tone gets more established. I used one ChatGPT chat for all of the copy I created for the main part of the site. This helped keep consistency throughout the pages.
I would not have been able to create the site as quickly as I did were it not for generative AI. Then again, if it wasn’t for generative AI, I wouldn’t have needed to make the site. It’s a bit of a chicken or egg situation.