GenAI for text and visuals: Marketing’s AI power couple

Can you recall a piece of content you were working on that was either all text or all imagery? As marketers, we often work with a mix of text and visuals — social posts, ebooks, blog posts, email campaigns, white papers, landing pages and paid advertising. 

Our best tactics combine words and images to help tell a memorable story, emotionally connect with our audience, and create meaning. 

For the most part, generative AI tools fall into two broad categories:

• Text-based tools like ChatGPT.
• Visual tools like Midjourney or DALL-E. 

While there is still some separation between these tool types, the growing trend is an integrated approach. This article examines the powerful combination of text-based and image-based generative AI.  

OpenAI: One and done?

If you’ve been following AI news, you know that major drama has unfolded at OpenAI. This upheaval led some to question the longevity of ChatGPT, OpenAI’s flagship platform, without Altman at the helm. 

While I may be conflicted about the ethical drama playing out in OpenAI’s board room, summed up nicely in The New York Times, the thought of not having access to ChatGPT and the other OpenAI platforms once again set off my AI anxiety alarm. 

Along with millions of other professionals, I have come to rely on ChatGPT to help boost productivity and creativity and do an endless list of helpful tasks. And while there are other AI models and platforms, OpenAI’s tools are among the best. 

They recently integrated DALL-E 3, their image creation tool, with ChatGPT and launched a functionality that allows ChatGPT to “see” and analyze images. This is a perfect example of how generative AI technology is starting to move toward a blended approach between text and visuals. 

Dig deeper: 6 ways to use generative AI for your marketing

My money is on Microsoft

Another example of this trend is the launch of Microsoft Copilot and Copilot for Microsoft 365. The biggest announcements during the Microsoft Ignite 2023 keynote were the enhancements to Copilot. Very soon, everyone will have access to integrated AI copilots across the entire suite of Microsoft products.

Currently available to enterprise customers via their early access program, Microsoft has reported huge boosts in productivity, with “77 percent of people who have used copilot saying they don’t want to give it up, 70 percent saying copilot makes them more productive and 68 percent saying it improved the quality of their work.” Microsoft is also working on rolling out Copilot to SMBs.

Microsoft, along with other big AI players like Google, Meta and Amazon, will continue to integrate AI tools into our everyday workflows and the platforms we frequently use. This may hurt the ecosystem of thousands of AI startups that have flourished over the last several months.

You can choose from dozens of tools that use generative AI to do specific tasks like editing videos, writing social copy, creating paid ad campaigns and more. While these tools may prove tempting, I have found that I always get better results working directly with large language models, like ChatGPT, DALL-E, and Midjourney.  

Marketing use cases for text and image-based genAI tools

Below are some thought-starters on how you can bring together AI text and visuals to help improve your campaigns. This is just a sampling of the ways I use these tools. 

• Create visual concepts for campaign imagery. You can copy and paste the campaign topic, description or any text-based information into ChatGPT and ask for 10/100/1000 visually engaging concepts or treatments. 
• Create specific Midjourney or DALL-E prompts. You can create text prompts for image generation tools using ChatGPT or DALLE-3. These tools can help you craft the best prompt possible in just a few seconds. 
• Get feedback on campaign visuals, logo treatments, product packaging and more. ChatGPT can now provide input on images. 
• Customer AMA. Ask your customer anything with your AI customer companion. You can create unlimited AI personas that provide detailed feedback on campaign topics, visuals, aesthetic preferences, values, motivations, etc. I use this whenever I’m working on a campaign and want to text my choices.
• Access a world-class creative director. You no longer have to pay a fortune to gain insight from an experienced creative director. Instead, you can instruct ChatGPT to provide feedback from the perspective of an award-winning creative director. 

Dig deeper: Third-party data in advertising: Best of the MarTechBot

A shortlist of AI tools to use daily

While I prefer to work directly with ChatGPT and DALL-E, there are a few third-party tools blending text and visuals that I find myself going back to repeatedly.

• Microsoft Designer: I love this tool to help create images with text overlay. It’s currently free and allows you to generate an image from text, select the aspect ratio, edit the text and download the finished result. 
• Canva: The popular tool now has an integrated image generation tool for its Magic Studio. My students love this tool as it allows them to easily comp together images with graphics and text. 
• PlayPlay: This tool helps you create videos using still images, video clips, text and animations in a flash. I created the video for my AI persona passion project, shown above, in just a few minutes. 
• GlossAI: This allows you to easily edit long-form videos, add captions and subtitles, text overlay and craft social short videos that include suggested social post copy. 
• ChatGPT: My one-stop-shop for all of my AI companion needs — whether it’s asking how to deepen my meditation practice, build a birdhouse or craft a campaign for my target persona. 
• DALL-E 3: One of the best image creation tools, it can manage adding text to images and now lives within the ChatGPT interface. 
• Midjourney: I still love Midjourney as it has many tools to help construct any image, from street-style photography to illustrations, cinematics and more. You can select specific film stocks and aspect ratios from hundreds and thousands of styles. 

Remember that all these tools are exponentially improving week to week. Today’s AI platforms are the worst version you will ever use. And while I may be making educated guesses on the tools that will stand the test of time, it’s impossible to predict the future.