Send us a message

Exploring the Opportunities Within Experimental AI Tools in Advertising

The first and most important thing I remind anyone who wants to discuss AI in the creative arena is: “AI” doesn’t stand for “artificial imaginations.” AI is intriguing and game changing as an executional tool set, but these tools expand and execute upon an original human idea. At least for now, imaginations still require human ingenuity and experience.

As one Writer’s Guild member recently penned on her protest sign: “ChatGPT doesn’t have childhood trauma.”

So, while AI won’t displace the most valuable aspect of creative talent, mastering it helps creatives be even more valuable by increasing the breadth and speed of crafting client-ready creative ideas. Savvy agencies will use this newfound efficiency to carve out more time for imagining on the front end, rather than just truncating already compressed timelines. Why? More time lets imaginations run and develop better-quality ideas.

Currently, our agency is using text-to-image mostly for rapid comping of key frames and storyboards — the heavy lifting to sell an idea. But AI is not only opening things up on the creative front. Some teams are using GPT platforms for optimized search ads, while others use them to identify popular songs for conceptual spots. And talent from all disciplines is using AI to improve and streamline their personal communications internally and externally.

Everyone at The Shipyard who wants to experiment with AI in their discipline is encouraged to do so, and the tools are made accessible. An experimental mindset is our litmus test for agency talent — that’s our jam. And having our media, data, strategy, production and creative teams all trying out AI in their disciplines means we’ll inevitably discover new ways to use the tools to integrate into our agency or create new value.

It’s also why we haven’t standardized our agency on any one tool set yet. Nor should we, given their rapid expansion.

Future state? The opportunities are huge. The Shipyard has long been an expert in machine learning in performance media. So, the expansion to AI-powered content and image generation aren’t entirely new muscles for us to flex. Our new creative content studio will be able to use AI for lower-fidelity, lower-cost executions in social. And the merging of machine learning performance media with AI-generated assets will be nirvana for on-demand optimization and personalization.

The biggest challenge to all of this is keeping pace with the inevitably evolving artistic copyright side of things. Two published authors just sued OpenAI because their digitally accessible works form the basis of “training” knowledge that ChatGPT is pulling from to create its responses (for which real humans are then taking credit). In simple terms, their published writings and ideas are being “sampled” without permission as part of how AI is writing its content. This will be an interesting case to watch, as will the hopeful resolution of the Writer’s Guild strike. AI will continue to be divisive and newsworthy because it challenges how we understand the world — but isn’t that what creativity is all about?

Originally Published by AdForum