Introduction
In a conversation with a friend, he mentioned using Midjourney to craft visuals for his luxury bag brand, Candamill. He explained that when he opted for an AI-driven approach in his campaign imagery, he contacted his usual collaborators. Instead of relying on their physical expertise, he proposed hiring them as consultants for prompt engineering. His reasoning was that their deep knowledge of their respective creative fields would enable them to describe intricacies effectively, resulting in refined campaign imagery. Can you guess their response?"
"No, we have no interest in artificial intelligence or participating."
Fast forward, Candamill drops their newest hand bag campaign and showcases a combination of generative ai and composite photography.
The Result: Not only did Candamill get more social engagement but also more sales.
My friend shared some impactful advice with me:
"At least play around just so you understand where AI is going. I'm a small business owner and if I'm already thinking about what jobs are going to stay, which will be phased out, and generally how creatives will have to adapt — I already know the bigger companies have thought about it."
- Prompt engineering enables users to shape the behavior and output of AI models. By strategically designing text based prompts, users can guide models to produce accurate, relevant, and context-aware responses for various tasks.
Tools I'm Currently Using
The landscape of artificial intelligence has evolved significantly in recent times. It's no longer limited to abstract discussions or the stuff of movie plots; instead, it's enhancing productivity and expanding the realm of artistic possibilities. Thanks to innovative applications like ChatGPT, Midjourney, DALL-E, Google's Bard, and even trusted tools like Photoshop, my perception of creative potential has transformed almost overnight. Generative AI is now apart of my creative process, offering a new avenue for exploration.
To clarify, while it's not flawless, this technology enables me to quickly test ideas without the financial burden. For instance, I've embraced Photoshop's new Generative Fill feature, which has become a literal magic wand in my image retouching workflow.
Additionally, Midjourney's blend prompt has empowered me to craft visual compositions and blend creative assets for inspiration in future image creation. This tool is particularly useful for generating mixed-genre moodboards.
ChatGPT plays a unique role in my creative toolkit, akin to having another "smart person in the room." I rely on it for tasks such as copywriting, critiquing residency proposals, and building resumes. While GPT's tone of voice is sometimes weird, it's helpful in establishing a starting point for a wide range of projects.
Similarly, Google's Bard, integrated into the Google app ecosystem, serves a purpose similar to GPT, providing valuable assistance in various projects.
Lastly, Spark's AI-driven grammar correction and tone enhancement prove especially beneficial in emails.
1+1 can sometimes equal 3
I'm sure some of you might be wondering, "This is fascinating, but I'm not quite sure how you're achieving these outcomes."
That's a valid question, and I've introduced this new generative AI image (above) with the purpose of shedding light on my process, approach, and the results I achieved.
My creative identity is deeply rooted in retro-futurism, and my pre-existing photographs help with Midjourney's "/blend" feature. Additionally, there are times when I draw inspiration from photos created by my favorite artists to craft unique cross over compositions. So, let's use these three images as our starting point within Midjourney:
We can see that Midjourney blended the visual qualities of each image in various ways:
1. Color from my personal work (image 1 and 2).
2. Created a swooping architectural backdrop from the rounded circle opening in image 1 and potentially the rounded highlight from image 2.
3. Attached 1950's-esque dresses from my images.
4. Emulated the refracting light in the shadow from Clarissa's photograph (Image 3).
5. The posing of the model in the pink dress looks somewhat familiar to the women in image 1.
6. Sharp shadowy diagonals definitely added from image 3 and potentially "weighted" as an important compositional feature due to diagonal/triangle features from all three images.
With small tweaks in photoshop, this is the final result:
I'm really pleased with the outcome! It captures the subtle qualities of my previous project work while drawing inspiration from the photographer, Clarissa Bonet. I can envision myself scouting photo locations in the future to create similar compositions using a traditional camera for higher quality results. This is where the benefit of AI lies — enhancing workflow and enabling playful experimentation.
AI should almost always be regarded as a tool, not the final result. Replicating and using another artist's work or resources, especially for commercial purposes, can raise potential legal and moral concerns that may not bode well for you or your clients. Until new laws are established for this emerging technology, it's wise to leverage these programs in unique ways. This could involve combining AI and photo assets for composite results, finding inspiration for composition moodboards, or even streamlining your creative process from 4 hours to just 1.
Be the creative director and concentrate on what truly matters. Remember, it's no longer about the tools; it's all about storytelling now.