OpenAI has introduced a groundbreaking upgrade to ChatGPT, enabling it to generate and modify images using its latest AI model, GPT-4o. CEO Sam Altman announced this enhancement during a recent livestream, marking the first major expansion of ChatGPT’s capabilities beyond text.
With this update, users can now create and edit visuals directly within ChatGPT and OpenAI’s video-generation platform, Sora. Initially, the feature is available to Pro users subscribed to the $200-a-month plan, but it will soon be rolled out to Plus and free-tier users, as well as developers using OpenAI’s API.
GPT-4o’s new image-generation abilities deliver higher-quality and more detailed visuals compared to its predecessor, DALL-E 3. It takes additional processing time to refine images, resulting in more accurate and realistic outputs. Users can also edit existing images, modifying elements such as backgrounds, objects, and fine details through a technique known as inpainting.
To develop this capability, OpenAI trained GPT-4o using publicly available data along with proprietary datasets from partnerships, including a collaboration with Shutterstock. While many AI companies keep their training data sources confidential, OpenAI has emphasized its commitment to respecting artists’ rights. The company has policies in place to prevent its AI from replicating the work of living artists.
Related Post
Brad Lightcap, OpenAI’s Chief Operating Officer, stated that the company provides an opt-out option for creators who do not want their content used in training. Additionally, OpenAI respects requests from websites that block its data-scraping bots, ensuring greater control over how data is used.
This latest enhancement marks a significant step for ChatGPT, positioning it as a more versatile AI tool that can assist with both text-based and visual creative tasks. As AI-generated content becomes more sophisticated, OpenAI continues to balance innovation with ethical considerations, providing users with powerful yet responsible AI tools.
