IN OUR HEADS
August 07, 2023
Sometimes you hit a wall. You need to come up with a new way to express the same key message or find a way to communicate a truly unique visual approach you can’t find in a stock image library. But you can’t find the right words or ways to bring what you see in your mind’s eye to life. Enter generative artificial intelligence like ChatGPT and Midjourney — tools that can provide the creative sledgehammer you need to break through.
I found that the content from generative AI creativity is an ideal pressure release valve to give your inner editor a break. I started a small experiment with ChatGPT. I asked it to provide 10 headlines and then another 25 headlines. The first 10 I’d already covered myself, but the next batch introduced new verbs I hadn’t thought of. Asking it to provide 50 headline options really made headway as more new verbs and adjectives were presented. I couldn’t cut and paste the lines directly into my copy deck, but it was enough to inspire a few approaches I hadn’t thought of. And it only took about 10 minutes to break through my creative block.
On the visual side, my team and I were trying to bring a concept to life with authentic, narrative imagery for a project. We wanted to find a way to get there without using straight-up stock imagery that would generalize the idea and pull it back into the realm of the expected.
Using Midjourney, my art partner began generating images with prompts based on the campaign’s film scripts. It was interesting to see what it created. Characters appeared as humanoid creatures with inexact features. Sometimes their features were combined in fantastical iterations, with lobster arms and melting landscapes as if painted by Salvador Dalí. Absolutely unusable. But through the process, we picked up approaches to composition and framing, providing direction that helped us find a way to unexpected visualizations.
We lifted nothing from the tools directly, but the odd visual compositions and fresh language cracked the work open, changed our perspective, and ultimately saved time.
The Potential Power and Pitfalls of AI With the Creative Process
One of the greatest potential impacts that generative AI tools like ChatGPT and Midjourney offer is saving time. AI and efficiency go hand in hand. And when it comes to the creative process, efficiency is difficult to maximize. Pushing harder against a creative block doesn’t speed up its solution. But generative AI is like a co-worker that is always available with nearly limitless ideas that can help you see what you are missing.
Sending AI on a copywriting exercise can help you accomplish an hour-long task in just a few minutes. For more junior writers, iterating and providing multiple options can be challenging. At the same time, part of the learning process is understanding what kind of line works and what doesn’t. Curating results from an AI prompt is a good way to help writers understand how to better evaluate their own work while getting time-saving assistance in developing variations.
But with any technology’s exciting potential comes the accompanying “buts.” This shows up as both AI bias and AI copyright issues. Creating generative AI is a bit like teaching someone something entirely new. If the teacher has an inherent bias towards race, gender, or other things, it will show up in teachings. And without any context of how a bias would be wrong, the bias is accepted and perpetuated. Examples of this include cultural appropriation in generated imagery and insensitive language.
And just as bias continues without context correcting it, AI doesn’t know what is and isn’t copyrighted. This makes it extremely important not to use content produced by generative AI as a final product. It’s up to the human operator of AI to ensure nothing problematic is leaking into the work.
How to Use Generative AI as a Creative Steward
AI’s issues are still in the process of being solved. Considering the concerns with ongoing development, the lack of guardrails, and the unknowns of AI’s capabilities, there’s a lot to be done for the future of AI.
In June 2023, the European Parliament began the process of establishing rules governing the in the EU. The proposed rules would ensure that AI aligns with EU rights and principles that seek to ensure human oversight, transparency, safety, privacy, nondiscrimination, environmental protection, and social well-being. It would ban facial recognition and other applications with serious risks to health and safety. It’s an important step forward and will hopefully provide a model for regulating the use of AI worldwide.
Until these rules are established, it’s up to the user to determine the risk of violating copyright as they use AI for creative work.
As a creative professional and someone who respects intellectual property rights, I think it’s acceptable to use generative AI as a source of inspiration and ideation — but not for a finished product. And for creative leaders, it’s important to see AI as a tool to augment the creative process. Through my experimentation and use of generative AI, I came to realize that it’s a way to help us rise above ingrained biases that seep into our thought processes not directly connected to ethnicity, background, education, or privilege. Like it or not, we all carry these kinds of biases, and that is why it’s essential to collaborate in diverse teams with a wide range of opinions and experiences. Like colleagues with different life experiences and backgrounds, AI provides a bias foil, and our teammates provide human checks and balances.
In the end, generative AI helps creatives transcend individual limitations. It’s a great tool for recharging the creative engine when teams run out of gas. It’s a brainstorming facilitator, a word finder, and a bias foil.
