Show and tell... and truth
- Hannah Vaughan Jones
- 2 days ago
- 2 min read
This week i've jumped on another seemingly viral trend and prompted #ChatGPT with the following:
"Create a caricature of me and my job based on everything you know about me"
Full disclosure - I did this twice, curious to see how far I could curate my own AI outcome. Both times I added three different photos for the image generation.
Here's the first:

Now, I did add some details about the work that I do - news anchor, podcast host, interviewer, Emmy-nominated journalist - so I can see how the various features showed up. I'm pretty sure I have that red dress (albeit not worn in a decade and not in the photos I prompted the AI with). The caricature does look a lot like me (the hair alone is spot on!). However, I have never interviewed Biden or Putin (why did these two feature as opposed to others, I wonder?).
I also wonder if this artificial intelligence "show and tell" production is in fact, false. Not just because it's an artificially produced caricature, but because its depiction of me is fundamentally inaccurate. Not entirely false, but certainly skewed. I am Emmy-nominated, which means I've not actually won one despite the suggestion in the image. My audience seems predominantly pale/male/stale... not accurate given most of my work in politics, health and energy is inherently global, hence multi-racial, cultural, diverse in socio-economic inclusion with gender equity at its core (hello #WomensHealth ). Anyway, it's got me thinking of the versions and productions of ourselves that we "put out there" to the world that aren't merely enhancements of the real thing, but potentially, misleadingly.... wrong.
For my second prompt, I merely added three photos without offering any context as to my job title or work experience...

This is MUCH more like it, I thought. In the pink, wearing an outfit I would wear any day (every day if it was school-run friendly), plus I look cute as a button.
I guess it just helped the AI that one of the pictures I submitted had my name on a screen behind me.
But the flowers and candles in the middle were actually from the Manchester Arena bombing atrocity of 2017. A time when yes I wore pink in respect to the 22 young lives lost that awful day. But I certainly wouldn't have been grinning like a cheshire cat like i'm depicted here in the foreground.
So, I guess what I'm trying to say is that a picture can tell a thousand words. But it doesn't mean those words are true.
In "showing" me in caricature form, the "telling" is misleading. And while no one was hurt in the making of this AI experiment, let this be a reminder that nothing is as it might first seem.