The pope in an immaculate down jacket, Emmanuel Macron in a demonstration… Lately, not a week has gone by without a new totally
fake
photo making the buzz on social networks.
Behind these fake images, bluffing but fabricated, we find so-called generative artificial intelligences (AI).
The most known ?
Dall-E (which reads “Dali”, pun intended), Stable Diffusion, or Midjourney.
From a simple text, called “prompt”, and with an almost childish simplicity, these new tools make it possible to create synthetic images.
If until recently the visuals produced lacked precision and remained easily traceable (especially in terms of the rendering of the hands), the new update of Midjourney, last March, marked a technological leap: the images are almost perfect.
In video, "We must regulate the practices of use of artificial intelligence", supports Aurélie Jean
Gender bias
On the occasion of International Women's Rights Day on March 8, the feminist and humanist movement #JamaisSansElles launched the campaign "Artificial intelligence maintains very real stereotypes" produced by the Havas Paris agency.
Havas Paris / Press
But lately, in addition to ethical issues, copyrights and manipulation of information, image-creating AIs are controversial…
This article is for subscribers only.
You have 83% left to discover.
Want to read more?
Unlock all items immediately.
TEST FOR €0.99
Already subscribed?
Login