I asked Bing’s Image Creator to create me an Easter themed Where’s Waldo, and, as it turns out, it’s not that hard to find Waldo.

It looks like Ai doesn’t quite understand what Where’s Waldo is. I thought I was providing a good prompt, but every image always had Waldo front and center.

Here’s what I started with:

A vibrant, hand-drawn Easter scene in “Where’s Waldo” style. A busy meadow filled with people hunting colorful eggs, playful bunnies, chicks, and lambs. Waldo, in his red-and-white striped shirt and hat, is hidden among the crowd. Cartoonish, cheerful, and packed with springtime details.

No Luck. It seems “is hidden among the crowd” wasn’t understood. Then I added “Waldo should be hard to find” at the end, and still no luck. I then updated it to say, “Waldo, in his red-and-white striped shirt and hat, is hidden among the crowd and hard to find” and still, Waldo is front and center.

The last version of the prompt, I updated it up to say “One person, in red-and-white striped shirt and hat, is hidden among the crowd and hard to find.” I was thinking that maybe if I don’t say “Waldo” specifically, maybe it’d catch on, but nope. Still got an image focused on Waldo.

This proves nothing, except that Ai image generation isn’t always as intelligent as some worry it is. Sure, it’s smart, and it cranked out Where’s Waldo style images in seconds, but it didn’t provide what I wanted. Even the images I did get had malformed people and animals. It’s impressive that it could understand what I was going for; too bad it missed the mark.

I’m sure if I were running my own Ai image generator, and had spent countless hours training it up on Where’s Waldo, I could output something great, but I’d rather just play with what’s out there today and see what it can create. Sometimes it’s amazing, sometimes it’s questionable. Either way, it’s fun.