This is an old project by Studio Moniker (Amsterdam)… I was once at a presentation of their work at Resonate in Belgrade (memories…) and they explained that the slight delay before the reveal of the underlying image is added artificially to add a bit of drama.

Btw, the correct image is loaded through a Voronoi diagram.

It's funny, im sure to Millennials this is jsut some nostalga to a simpler internet but I find seeing old dorm rooms and random slices of life from more than a decade ago really intresting, from the fashion to the red eyes from the flash. half of these photos look like they could appear on a modern Lo-Fi album cover

I didn't like that there was a long delay after you move your mouse, and since all the calculation happens on the frontend, i figured you could easily make it ~instantly update and follow you around. So i (well, claude) made a version that does that and i think it's more fun. here's the code you can paste into the console at pointerpointer.com: https://pastebin.com/f7YqQNxg

Along similar lines:

> On May 4th, 2007, we asked internet users to help isolate Michael Jackson's white glove in all 10,060 frames of his nationally televised landmark performance of Billy Jean. 72 hours later 125,000 gloves had been located. wgt_data_v1.txt (listed below) is the culmination of data collected.

— https://www.whiteglovetracking.com

Since this is old, I assume someone found these photos and then manually selected the pointer location. Maybe used openCV or something like that.. But I'd most likely go with manual.

There are 700+ images defined in https://pointerpointer.com/new-positions.json and the script finds the closest match to the current mouse pointer.

This is hilarious. I thought it was gonna be one of those pointer animation nightmares, and I was going to regret clicking on it, but I still kind of wanted to click on it anyway. But actually, it’s really funny and amazing to think about how you created it.

This one is an old one. I still love it, now with nostalgia. I always wondered how they got all those pictures. It looks like there is a recurrent theme tho.

Paul Irish explained the basics of how it works here: https://www.youtube.com/watch?v=Z2ZXW2HBLPM

I wonder if you could create something similar nowadays using generative AI but with the finger in a very specific location without pregenerating thousands of images.

Can you make a version with cats in it? Like cats jumping at the cursor or trying to catch the cursor with their paws.

How does it work? Does it contain all the possible images with al the possible pointer positions? Does it do some corrections, such as rotation or shift to the original image?

I'm surprised how well this works on mobile. I thought the photos would be distorted at least.

How do they acquire these images?

This is apparently from 2012, but I could've sworn it was older than that.

I think this was created by the people at the art collective of monochrom.at,

This is where I have seen this the first time ca. 2006.

Edit: Now that I searched for it I only found their project ZeigerPointer, where they collect such images, maybe I mixed it up.

good old web2 fun