To make sure you never miss out on your favourite NEW stories, we're happy to send you some reminders

Click 'OK' then 'Allow' to enable notifications

AI creates what the average person with a common first name looks like

AI creates what the average person with a common first name looks like

Do you agree with the results?

Since hitting the mainstream through platforms like ChatGPT, artificial intelligence (AI) has been really put through its paces.

Tasked with creating words and images with accuracy that would probably take humans several days to match, the novel tech has become something of a fascination for many.

But it didn’t take people long to use it for laughs, and one of those trends involves asking the image-generating AI platform - Midjourney - to visualize what people might look like based on different criteria. In the last few weeks, we’ve seen it conjuring images of what AI thinks the average couple looks like from every country, or what AI thinks you look like based on what dog breed you own.

We totally buy AI Anthony - but it depends on what kind of Anthonys you've come across.

Well, it's at it again, this time with a new lens of what the average person with a common first name looks like.

There are 10 characters in total with the names Anthony, Charlotte, David, Emma, Henry, Joseph, Mia, Michael, Olivia and Sophia.

Charlotte looks about right to us.

While Anthony and Charlotte look like probably every Anthony and Charlotte we’ve ever met, the image for David is a little off-kilter, should we say.

Do you agree with AI's interpretation of someone called David?

While it totally depends on your own experiences of who you’ve met, the first image that comes to mind when thinking of a David, for us, isn’t an arty and well-dressed middle-aged, grey-haired man.

Yep, he pretty much looks like every Henry ever.

Henry, on the other hand, is a particularly funny one, as he looks just like you’d expect a Henry to look.

Mia looks like a cool person - you might be flattered if you share the same name.

The same for the AI-generated Mia.

This Michael might prove divisive.

As for the Michael, we’re not too sure. It doesn’t look like the Michael we’d think of when hearing the name.

Take a look for yourself, though - you might not feel the same way. There are also representations of the 'classic' Sophia, Emma, Olivia and Joseph.

Is anyone else getting Emma Stone from this Emma?
We're not sure why all the women are young, but this Sophia seems about right.
Joseph's all-beige look is actually quite chic.
It's a bit hard to focus on Olivia, what with her distracting hat.

The outcomes of this AI experiment are certainly interesting, but nothing as crazy as the AI-generated images of what Europeans supposedly think Americans from each of the 50 states look like.

Creating quite a commotion online and even offending some people, the states featured in this experiment included Texas, California, Colorado, Florida, Oregon and Maine. A number of Reddit users criticized the choice of prompts used to make the images - especially for the “red states” - while others from Maine humorously requested a stop to the lobster-themed depictions.

Texas was perhaps the most controversial, with the AI portraying a man in typical 'cowboy' attire, complete with a large hat, brown shirt, blue jeans, and a substantial belt buckle.

In California, the AI-created human epitomized flower power, sporting long hair, big sunglasses and a floral shirt. Colorado, on the other hand, depicted a woman amid a marijuana-filled mountaintop, donning a green hoodie and headband with what appears to be a joint in her hand.

Meanwhile, in Florida, the AI envisioned a man with a long white beard, dressed in blue shorts, a baggy pink shirt and a sunhat.

While these images are good for a quick laugh, It's important to remember, of course, that there's no real “average person”, but it’s the stereotypes, fashion trends and local customs that exist which seem to have influenced AI's creative process in generating them.

Featured Image Credit: Credit:

Choose your content: