Phylo - Learn how to sequence a genome!
A game to enter the fascinating world of genetics.
Publish at December 10 2025 Updated December 10 2025
We're at a turning point where generative images are asserting their presence. A.I. is capable of creating images, music and videos that are nothing short of realistic. Yet anyone interested in the phenomenon realizes one thing: there's something disturbing about these images created by algorithms. Something inhuman. And with good reason, as this very artistic ARTE report shows, the singularity is disappearing.
Artificial intelligences don't create. You'd think they would, but in fact they only reproduce. As information scientist Philip Di Salvo puts it so well, they count. That's all they know how to do. Calculate statistically according to our requests to offer an answer, a close creation, without imagining. So they start from a pre-existing framework that they can't question. So, if the framework is sexist, racist or based on patriarchal aspects, everything they come up with will be tinged with that.
What's frightening for the future of AI, then, is the loss of singularity. The images designed show a smoother side, which fails to reproduce the human shadows, the differences, those little features that make us what we are. All the more so as we'll soon reach a point - and we're probably there already - where AIs will no longer be fed solely from human sources, but also from old calculations by the algorithm or others.
Standardization lies in wait. It's up to creators, artists and us humans to make sure that, alongside AI, there are real productions with imperfections and unique aspects.
Running time: 25min33
Image: JetalProduções from Pixabay