This is how I work with Gen AI as my partner to bring my ideas to life:
- First, I have an idea. (A ninja cat)
- I ask Gen AI to generate multiple versions of my idea. (“Midjourney, give me different ninja cats”)
- I select and shape the idea. (I refer to Midjourney’s ninja cats and use my hand to sketch out a ninja cat using my iPad.)
- Gen AI generates a better version of the idea. (“Midjourney, take my sketch, and here’s my prompt to make it into a painting.”)
- I iterate with Gen AI. (I try different prompts until I get what I want from Midjourney.)
- I manually edit Gen AI’s work to my liking and finalize it. (I edit the visual in Photoshop to finalize my vision.)
At each point, I call the shots using my prompts. I cannot get too hung up when Gen AI doesn’t give me exactly what I want (eg. the differing expressions of the ninja cat). I have found that Gen AI is like a human who insists on having some artistic licence. As long as my original idea remains intact, I’m good with it.
This is how you should work with Gen AI as well. Wield its vast power, go back forth with it, and don’t give up your control or your humanity.