Generative AI for Inspiration

Life Science Marketing Radio

David Nathan wrote a children’s book in the carpool line at his kids’ school. ChatGPT delivered the text and he used Midjourney to illustrate the main character.

Generative AI applications like ChatGPT and Midjourney are dominating conversations in the marketing world recently. The last episode of cc: Life Science covered possible uses and concerns around Chat GPT which generates text responses to user prompts. Midjourney is a tool that will generate images based on your prompt, having been trained on a gazillion JPEGs from who knows where.

David is an agency owner diving into Midjourney to explore it’s potential uses in marketing. We talked about where he sees value as well as some of the limitations.

As an agency working with designers and clients, presenting concepts and getting feedback can be expensive even before you get to the deliverables. And not everyone is gifted with the ability to express an idea through a rough sketch, let alone words. David sees a benefit in being able to present a basic idea of a trade show booth, for example, and get feedback from the client before giving the task of detailed design to a professional.

I'm always finding myself in a situation where I have ideas and then I have to try to translate those ideas into, like understandable language for design. to have them then create the ideas. So not only can I now create some things myself at a pretty high level… you would have to be really, really talented digital renderer to create some of these images. But I can also use that as a starting point to help me communicate my ideas to the artists now too, right?

One of my (many) concerns is that, while you can prompt for images in a certain style, everything I’ve seen has a similar vibe to it. I don’t know if that’s because I know it came from an AI or what. David says that each platform has a certain “lean” to it. Will we become numb to seeing a universe of media that all feels like it was generated by a machine has no soul?

What does this do to artists and creators? One consideration is whether they should be compensated for the use of their work in training the algorithm. The other is simply their continued employment. Having said that, photography didn’t eliminate the art of painting. One could argue that it opened up new possibilities for painters to explore different ideas as opposed to reproducing reality for royalty.

Finally, we talked about the children’s book he made in the carpool line. David noted that ChatGPT came up with a reasonably good effort based on his prompt. And Midjourney was able to present a character that satisfied the imaginations of the five kids in the back seat (a monster like something you might find in a Pixar movie.)

What was interesting to me is that you couldn’t illustrate the whole book this way. For example, “Now show the same monster doing X.” Although the AI creates a “monster” image based on its training from other monster images, it doesn’t really know what part of the image is a monster. I found this fascinating and, honestly, inspiring in another way. Human beings are still amazing.

My hope is that we can use AI thoughtfully while conscious of not losing our humanity.

What are your thoughts? Have you tried Midjourney or Chat GPT? Leave a comment below.

Schedule a 15-minute chat with Chris about turning conversations into content for your life science company.

Intro Music stefsax / CC BY 2.5

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit cclifescience.substack.com