While navigating the maze of my web redesign and sharpening my service offerings, I found myself at the crossroads of developing customer personas. I have not had a lot of experience in this space, so naturally, I turned to my trusty AI assistant, ChatGPT, for a helping hand.
Feeding it with the information I had (and let's be real, it wasn't a lot), ChatGPT churned out three intriguing personas: AI Curious Catherine, Chamber Chris, and Educator Ellen. I was pleasantly surprised by the depth of the descriptions, and if you're curious, you can sneak a peek and click Personas.
Feeling adventurous, I decided to take things up a notch. Using the ChatGPT-4 and DALL·E combo, I requested images to accompany these personas. And that's when things got...interesting.
Anything stand out to you?
All three images had a common thread — they all showcased white individuals.
These biases aren't the AI being mean-spirited. It's echoing back the biases in the data it was trained on. And while some might shrug these off as minor hiccups, it's important to remember they can shape perceptions, steer decisions, and can even reinforce stereotypes.
I had to quench my burning curiosity, so I popped the question to ChatGPT: "What's the deal with these biases?" Its response was candid, acknowledging the bias and emphasizing the goldmine that user feedback is. But it's not just about the feedback. It's about the perspectives. What's evident to me might be invisible to you, and what you spot might be completely missed by me. It's the diversity of the human perspectives that matter most when the data is biased. I recently asked ChatGPT for some images of a doctor delivering a baby. I was pleased with the results, but that's just my vantage point.
So, If you've been side-eying AI, wary of its biases, you obviously are not alone in your concerns. Tomorrow, President Biden is expecting to announce an artificial intelligence executive order. The highlight? Advanced AI models would need assessments before federal workers can use them. Plus, there's a push to make it easier for highly skilled tech workers to immigrate to the U.S. For years, Silicon Valley has been pushing Washington to lower the barriers for their much-needed engineers and other tech geniuses. It's apparently a mad dash for technical talent, and America wants to be in the lead. And it doesn't stop there. The State Department is also gearing up to launch a fresh program that's all about attracting the best in artificial intelligence talent. Stay tuned!
Now, with all this buzz at the national level, you might wonder, "Where do I fit in?"
Here's the thing: we don't have to be in the White House, Congress, or a tech company to make a mark. My challenge to you is rather than just observing from the sidelines, get your hands dirty and proactively engage with generative AI tools like ChatGPT. Understand its workings, question its outputs, and be part of the broader conversation.
Based on what my trusted source, ChatGPT told me tonight, every observation, every piece of feedback helps in shaping a more equitable digital realm. And for now, I'm going to bed believing it.
It's a collective endeavor. Are you in?
And one final AI generated image to close the blog. Here's the prompt I put in, "Could you help generate an image for the end of the blog showing a group of people smiling... with thumbs up... it will be the response to the question "are you in?"
Your turn to be the judge. What do you think? Please comment with your thoughts and don't forget to like and share the blog. It's very much appreciated.
Comentarios