- October 21, 2022
- AI Projects
When, at the beginning of September, we started designing this special issue, DALL·E 2 – the most famous software for generating images by artificial intelligence – was only available on a waiting list. For a few weeks, it has been open to everyone, and already the digital networks are flooded with more or less realistic images of cosmonaut kittens à la Picasso and elephants in tutu version of heroic fantasy. Stupid gadget? Make no mistake: these platforms are not (only) the new digital toy that serves to procrastinate in the office. Their intuitive access based on natural language, their overwhelming power (you only really measure it by trying it yourself) and their massive democratization place us on the threshold of a new world. An image revolution is underway, at least as important as the invention of photography (which, let’s remember, did not kill painting, but broadened its perspectives).
As tributary media and image specifiers, we very early had the desire to wrestle with this new power. Thus, for this special issue, we have teamed up with Mathieu Bernard-Reymond, an artist from Lausanne who is already familiar with these tools. For us, he generated almost all the images in this edition (only one, in these pages, was drawn by a flesh-and-blood illustrator. Can you recognize it?).
An image generator cannot do everything
From this exceptional collaboration, we learned three lessons. First, that chance must be embraced. The current – and temporary – imperfection of these tools harbors tremendous aesthetic and humorous potential: let’s take advantage of it! On the cover of this edition, we have chosen not an image created from a formal request (the attempts we have made in this direction have proved strangely disappointing), but an “accidental” image, the fortuitous product of attempts hazardous, an image that no human being had imagined. We could have relegated it to the blooper, we made it the front page.
Second lesson: we quickly understood that an image generator cannot do everything. For the time being, the possibilities of representing existing personalities or specific places remain limited (but this is progressing very quickly). This is why we have carefully avoided including in this summary interviews or reports that would have required documentary iconography. We do not do this kind of special issue without some editorial contortions.
Third lesson – the most important: the work of generating images cannot be entrusted to just anyone. Only someone who is experienced in producing, editing, and curating images is able to find the right solutions in the bewildering abundance that these platforms offer.
So no, press publishers will not be able to do without their photographers or their iconic services tomorrow. On the other hand, it is obvious that the work of the latter is set to change in depth. Each technological advance carries its share of fears. What jobs, what work is this software threatening? Those of the artists? Certainly not. They and they will appropriate (or not) this new medium, which will increase their practice, diversify it, shake it up, no doubt for the better (this issue’s portfolio is a brilliant demonstration of this). On the other hand, the legitimate fears are on the side of the small-time “creatives”, all the makers of mediocre images who lay out to-order pretext illustrations and ugly advertising campaigns.
Like all important technical advances, the generation of images will redistribute the cards: business models will die and others will rise from their ashes. Like all revolutions, this one will benefit those who, through use and reflection, will understand its potential and limits early on. We at T magazine have just begun this apprenticeship.
- Google presents the Gemini project, which sharpens the global competition in the AI
- AI-Powered Image Creation Contributes to Carbon Emissions
- Understanding Google Translate
- Amazon introduces Q, a cutting-edge business chatbot powered by generative artificial intelligence
- PyTorch vs. TensorFlow Frameworks
- Scientists Create Artificial Intelligence Model for Forecasting Stock Market Movements
- GitLab Improves AI Offerings with Duo Chat
- DeepMind’s System Delivers 10-Day Weather Predictions in Just One Minute
- AI Technology Empower Users to Choose Their Preferred Sounds in Noise-canceling Headphone
- Recommendation Algorithms
Get regular updates on data science, artificial intelligence, machine