Those who have the eye will have noticed that a large part of my posts here are illustrated by images produced by an artificial intelligence, in this case Midjourney.
For input, I compose a short sentence in English to describe what I want, often accompanied by the terms "vector illustration" for a drawing or "photo-realistic" for a photographic result. And the program outputs four breathtaking proposals that I can transform or expand. All for 8 euros per month.
For example, here are the four suggested results for "French Baker shooting films with a camera and a baguette":
Which is not to say that graphic designers have become useless - not at all! They can still bring coffee or do the short ladder to help catch high objects.
Just kidding (embarrassed emoji).
If the images generated by A.I. are always of surprising quality, it is still very difficult to have real control over the result. The machine does what it wants. Not to mention the supernumerary members and other graphic aberrations that we do not fail to notice at second glance. Smart illustrators will therefore make these tools their allies by composing the right requests upstream and making the right corrections downstream in order to create richer works, faster. Or so we hope.
But yes: these jobs are evolving. Along many others. Ask GPT3.
