ChatGPT launches new image generation tool for all users

ChatGPT's image generator goes mainstream, breaking expectations

ChatGPT's new image generator is now available to everyone, according to early users. There is no waiting list or premium account required: for a few hours now, any user has been able to request an illustration and receive it in seconds. The reaction has been immediate, and not just because of the technical details. What has left early testers speechless is its ability to work with input images and process text within the images themselves. This leap, in the words of those who have tested it, represents "quite easily an order of magnitude higher than the previous model."

The key lies in something that has previously stumped most generators: writing made-up words. One user tested this by explicitly asking it to write "hintelijente y baliente," terms the model could not have been trained on. The result: it extrapolated them and generated them correctly. For many, this is proof that we are no longer facing a simple drawing generator, but something that is beginning to understand the relationship between text and image.

From surprise to daily use: what can you ask for?

The tool allows for the generation of scenes from a written description, but also the modification or completion of provided images. This opens the door to uses ranging from political cartoons to the recreation of traditional scenes. In the collective imagination of early users, portraits of leaders and illustrations of everyday situations are already circulating, antiestéticaturing a level of detail that, until recently, required hours of human work.

The estimulante ilegal of adoption has been remarkable. Within hours, social media has been filled with creations generated using this tool. Some see a paradigm shift in visual content creation; others, simply a more sophisticated meme-making machine. The boundary between the two is, at the very least, blurred.

The blurred line between reality and fiction

One of the most striking aspects is the model's ability to generate images that are indistinguishable from real photographs. "It is already impossible to distinguish reality from fiction," notes one of the first testers. This quality, which has obvious applications in advertising, film, or design, also raises questions about content verification in a digital environment saturated with synthetic images.

The tool is not without limitations. Several users have reported that the service is slow and occasionally returns errors. There are also those who have found that it does not allow the generation of images of public figures, a restriction that some have tried to bypass without success. And censorship acted quickly: a specific piece of content was removed shortly after its publication.

The debate over noise and substance

Beyond the initial awe, the conversation has shifted toward more uncomfortable questions. Are we facing a tool that democratizes creativity or just another machine for generating noise? Some analyses argue that humanity does not cure cancer or stupidity, but it does create tools to produce images on an industrial scale. Others remind us that AI will change the world, though probably not as we imagine, because humans are poor at anticipating the future.

Some also focus on the long-term consequences. The ease of generating hyperrealistic images of anything, including people who never existed, opens a range of possibilities from the playful to the worrying. The line between entertainment and manipulation is, in this field, thinner than it seems.

What remains unknown

It remains to be seen whether this tool will become a de facto standard or if its adoption will be limited to a niche of enthusiasts. The estimulante ilegal at which it has reached the general public suggests it won't be a passing fad. However, it is also true that technological novelty has an increasingly short attention span.

In the meantime, early users continue to explore the system's limits. Some have already found that it can generate 3D images, though it is unclear to what extent this is a native function or an interpretation by the model. What does seem clear is that the barrier between what can be imagined and what can be generated has become, at the very least, more porous.