Chatgpt now creates images with context. And Mems – computers

Are they just Memes? What is the idea Chatgpt Create more context images, but Mems are one of the example of the probability of new activity Opena He has now announcedAnd Sam Altman brings the UAU effect I have not experienced for some time.

In the Image Generator Presentation Live Stream, which is now GPT – 4O, Opena CEO said that integrating the ability to create images in their LLM models is one of the objectives of the goalAnd the pictures are not only beautiful, but also useful.

To show an instance, Sam Altman took a selfie with his performance mates. The film has been included in Chatgpiti and requested to be transformed. It took some time, during this time the dates were displayed, but the result was directly displayed.

See the examples of the images created

Openai wants it to be tGiving conversation with GPT, explaining what you want and creating images including specifications such as size, colors or backgrounds. But it takes time and wants some patience. “The model has created more detailed images and it takes some time to render it, sometimes more than a minute,” he explained.

Openai has already introduced Sora for the video, and the creation of the image Doll-E (already in Generation 3), but still worked separately. This is another step to have a real multimodal model, as the CEO of the company is emphasized.

This performance is not certainly alienated, Elon Musk is sharing the last two new image editing features Your productive AI model, grocer. The race that adds more features to LLM models has been renamed and there seems to be news on behalf of various companies every day, which tries to obtain space in chatbots market and productive artificial intelligence agents.

Create images based on text and context

Openai describes it This model has been trained in the joint distribution of images and text and learn not only in the way the images are associated with the language but also among themselves. Pictures created as a result of post-workout work are consistent and sensitive to the context

Watch the video shared by Opena

With integration in Chatgt, It is possible to edit images throughout the conversation following the context of what happens. And the Openai guarantees coherence in the consumer’s interactions, as an example of the creation of a character from the video game, which is the detective cat.

Sam Altman confesses that there are still some limitations and the model is not perfect, but it is developing. And it ensures that it does not want to limit creative freedom, but obstructions are applied to the elements that violate safety rules. And transparency is also important, so all the images have C2PA Metadata and there is a motor that allows you to verify that some content is produced by the GPT-4O model.

Sexual harassment on children, sexual dipfakes and nudity or graphic violence are part of the type of content.

The new activity is now available for GPT-4O Plus, Pro, Team and Free Access and Sora, but also at Doll-E GPT. Soon the Enterprise and EDU models will be reached and developers can make images through API in the coming weeks.

Writing Note: The news during the show was updated with more information and images. Last update 19H02