show index hide index
- With the introduction of native image generation within GPT-4o, even free users have access to substantial creative power, allowing them to generate up to three images per day.
- At the same time, the question of the ecological footprint of artificial intelligence is emerging, fueled by the immense energy consumption of GPUs required to generate images via GPT-40. This paradox between accessibility and sustainability poses a significant challenge that the field of AI will absolutely have to resolve.
Discover the astonishing evolution of images generated by ChatGPT in just one year. What was once a simple geometric representation has quickly given way to stunning photorealistic creations. Dive into this remarkable transformation, which reflects the dazzling advances in artificial intelligence, now capable of producing works of unprecedented quality, defying user expectations and redefining the standards of visual creation. In just one year, the evolution of ChatGPT’s image generation capabilities is both impressive and fascinating. What were once rudimentary creations have now become a veritable playground for creativity, thanks to the advent of GPT-40. Users can now generate photorealistic masterpieces from simple queries. Let’s discover this extraordinary transformation. Humble Beginnings, Tenfold PowerIn 2024, when ChatGPT was asked to draw a bird on its free version, the result was rudimentary, to say the least. A simple image created from basic geometric shapes, composed of triangles and circles, produced an almost childlike rendering. This minimalism undeniably limited users’ artistic potential. A year later, this same request now generates a photorealistic work, packed with detail and reflecting a more refined visual understanding. This qualitative leap is explained by the incorporation of GPT-40, which has revolutionized not only image generation but also our perception of the capabilities of artificial intelligence.Unparalleled Accessibility for AllPreviously, free users of ChatGPT were limited to simple images produced with DALL·E 3, subject to a limited quota of two creations per day. The renderings, while interesting, often lacked
Artistic finesse.
With the introduction of native image generation within GPT-4o, even free users have access to substantial creative power, allowing them to generate up to three images per day.
This change has paved the way for stunning renderings, where images are studded with realistic textures, well-balanced visual compositions, and even the integration of readable text. The ability to understand complex and rich prompts has also improved, giving users the ability to create hyper-realistic action figures or fantasy worlds worthy of video games.Viral creativity that disrupts the norm
The rise of ChatGPT has generated unprecedented enthusiasm. Faced with Midjourney, the pioneer of photorealism, ChatGPT has been able to make up for lost time with a more fluid and accessible image generation experience. In March 2025, this achievement generated considerable media hype, so much so that the massive influx of users caused OpenAI’s servers to overload. However, this success does not come without controversy. The reproduction of iconic artistic styles, such as those of the famous Studio Ghibli, has rekindled debates around intellectual property. The generated images, evoking Miyazaki’s dreamlike universe, raise questions about the legitimacy of using protected references without explicit consent.A burgeoning ethical debate
At the same time, the question of the ecological footprint of artificial intelligence is emerging, fueled by the immense energy consumption of GPUs required to generate images via GPT-40. This paradox between accessibility and sustainability poses a significant challenge that the field of AI will absolutely have to resolve.
Although visual creation is more accessible than ever, the gap between free and paid users remains. ChatGPT Plus members enjoy a more powerful experience, thus exacerbating the inequalities in access to this graphics revolution. As we approach 2026, what developments can we expect to see in this rapidly expanding field? The emergence of automatically generated videos or immersive augmented reality experiences could well mark the next steps in this adventure. ChatGPT’s visual boom is just the beginning of an odyssey that promises to redefine our creative universe.
To read Giorgia Meloni : quand l’intelligence artificielle crée des images surprenantes en lingerie
