Blog Post

Prmagazine > News > News > ChatGPT’s image-generation feature gets an upgrade | TechCrunch
ChatGPT’s image-generation feature gets an upgrade | TechCrunch

ChatGPT’s image-generation feature gets an upgrade | TechCrunch

On Tuesday’s live broadcast, Openai CEO Sam Altman announced the first major upgrade chatgpt More than a year of image generation ability.

Chatgpt can now take advantage of the company’s GPT-4O Models create and modify images and photos locally. GPT-4O has long supported AI-powered chatbot platforms, but until now, the model is able to generate and edit text instead of images.

Altman said GPT-4O local image generation live today in OpenAI’s AI video generation products Chatgpt and Sora, for the company’s $200 Pro plan for $200 per month. Openai said the feature will be available soon, as well as Chatgpt’s free users, as well as developers using the company’s API services.

GPT-4O with image output “thinking” is longer than an efficient alternative image generation model, dall-e 3make the content described by Openai more accurate and detailed. GPT-4O can edit existing images, including those used with the people in it – converting them into foreground and background objects (such as foreground and background objects).

Openai did not disclose the image data it used to enable the new image generation feature. Many generated AI vendors view training data as a competitive advantage, so keep it in touch with any information that is close to the chest. However, training data details are also a potential source of IP-related litigation, another inhibition that companies reveal a lot.

OpenAI provides an opt-out form that allows creators to request that their work be deleted from their training dataset. The company also said it respects the request to prohibit its online clipping robot from collecting training data (including images) from the website.

Chatgpt’s upgraded image generation feature follows Google’s Gemini 2.0 Flash’s experimental native image output, one of the company’s flagship models. Powerful features spread on social media, but not necessarily for the best reasons. Gemini 2.0 Flash Image Component Results Very few guardrailsallowing people to delete watermarks and create images that depict the characters affected by the version.

Source link

Leave a comment

Your email address will not be published. Required fields are marked *

star360feedback