< IMG SRC = "/Uploads/Blogs/AB/B5/IB-FR8O5R46T_DF4A499E.jpg" Alt = "New Function: Chatgpt users can now generate images"/> ~ < P > Openai has announced that Sora and Chatgpt OB & Rsquo; Rsquo; join the efforts to generate images, about this & nbsp; nbsp; writer; nbsp; tech crunch. We tell what is known about the new feature. & Amp; nbsp;
< h2 class = "wp-block-glading" > that happened
< P > CEO OPENAI Sam Altman announced the first major renewal of Chatgpt image generation. Chat bot can now use the model & nbsp; GPT-4O & nbsp; to create and edit images and photos. The GPT-4O has long been implemented to the chatbot platform, but this model could still create and edit only text, not image. Altman said that GPT-4O image creation is already available for some users in Chatgpt and Sora. & Amp; nbsp;
< p > sora & mdash; It is also an OpenAi product to create artificial intelligence videos for the company's subscribers in PRO for $ 200 per month. Ability to create videos and images will soon become available for Plus users and free Chatgpt users, as well as developers who use API service.
< p >GPT-4O, although slightly slower than Dall-E 3 (another assistant & nbsp; companies to generate images), but it provides creation of more accurate and detailed images. Also Nofa function allows:
< Ul > < li > edit the image;
< li > to change and complement the details;
< P > This feature was made possible by learning a model on a large amount of public and private data received from partners such as Shutterstock.
~ < h2 class = "wp-block-glading" > why is it interesting< p > Generative artificial intelligence suppliers, such as Openai, are faced with a dilemma: on the one hand, training data is their competitive advantage, so they try to keep them secret and on the other & mdash; The same data can cause intellectual property lawsuits. Companies such as Openai try to find a balance by offering forms of refusal for creators whose works are used in educational models and declare respect for the rights of artists, prohibiting direct imitation of their works.
~ ~ < P > Updated the Chatgpt image generation feature, which is analogous to Google Gemini 2.0 Flash Experimental Function, has become popular on social networks, but not without problems. Users have found that the Gemini 2.0 Flash system has vulnerability that allow you to delete watermarks and create images of copyright -protected characters.