The e-paper news feed illustrates headlines with AI-generated images

It’s hard to read the headlines today without feeling like the world can’t get any worse. And then tomorrow rolls around and a new set of headlines disprove that thought. On a macro level, there’s not much you can do about it, but on an individual level, having your news feed show mostly inaccurate, AI-generated images can make things a bit off.

Let’s explain. (Roy van der Veen) liked the idea of ​​an e-paper display newsfeed, but the crushing weight of the headlines was a bit too much to bear. To lighten things up, he decided to use static diffusion to illustrate his feed, displaying both the headline and the generated image on a 7.3″ inky 7-color e-paper display. Every five hours, a script running on a Raspberry Pi Zero 2W fetches a headline from a random source — we’re glad the list includes Hackaday — and adds randomly chosen prefixes and suffixes, creating a prompt for a static spread based on the headline. To spice things up. For example, the prompt might look like, “Gothic painting off (driving a motor with an audio amp chip). Sad, dramatic, wonderful, dreamy.” You can imagine the results.

We have to say that, from the examples (Roy) shows, the idea works to a great extent — sometimes the images are so far away that it’s enough to figure out how the steady spread came with them. We would have preferred if the news of the Libyan floods had been buffered by a slightly less depressing scene, but it was certainly amusing to discover that what was supposed to be a “ritual mass murder” was really just a yoga class.

Leave a Comment