What Biases Do You Bring to Your Projects
This Week In Writing, we explore biases in our creative pursuits and how those biases can translate to AI-generated content.
Last week I was discussing a nonprofit client with a colleague. The client received a large check from a government agency, but they were contemplating moving away from accepting government grants. I asked, “why would they want to give that up?”
For most small organizations I work with, jumping through the hoops to receive significant government funding is something to be celebrated, not rejected. That’s the bias I brought to the conversation. In hindsight and more discussion, there are many reasons the organization might want to abandon government grants; namely, the restrictions imposed on receiving funds might be outside the organization’s mission.
We bring biases everywhere we go. In my job working with nonprofits, I must realize when my biases may filter a conversation with a client. As writers, we must also acknowledge when our biases may shape how we share our ideas with readers.
I was thinking about this situation while playing with Midjourney, one of the popular AI image generators. I read that when the AI doesn’t know how to interpret the prompt, it often defaults to an attractive young woman in some form of decay. Typing that out makes it sound horrific, which is why the bias here is so puzzling.
Why is AI biased toward making attractive women melt or peel into dust? Why is AI biased toward defaulting to beautiful young women in the first place?
I tested the theory by prompting test and 8675309 — I guess we can call the various decaying women in the images Jenny. Leaving the AI model to basic interpretation with these simple prompts exposes the system’s bias.
The thing about biases is that they aren’t necessarily good or bad, they just are — granted, if you’re biased against people because of who they are then your bias is bad, but that’s for a different discussion. We need to be aware of our biases and how they might influence us in ways we don’t realize or don’t intend. Whether it’s when working with clients, our writing, or relying on AI content creation, biases often affect the outcome.
How do you identify the biases in your writing and creative projects?