The artificial intelligence firm OpenAI unveiled an updated version of its DALL-E picture generator. It integrated it into its well-known chatbot ChatGPT. The company is sponsored by Microsoft. It is making the move in an effort to preserve its leadership in the new field of generative artificial intelligence in the face of growing competition. From established players like Google and upstarts like Midjourney and Stability AI.
ChatGPT and DALL-E 3 Combination
The picture generator, dubbed DALL-E 3, according to OpenAI, can comprehend a great deal more complexity and depth than the company’s earlier systems. Particularly complicated cues, to produce “detailed and accurate” images. The Sam Altman-led company claimed that. “DALL-E 3 represents a leap forward in our ability to generate images that exactly adhere to the text you provide.” According to the firm, the new image generator significantly outperforms the DALL-E 2 predecessor in terms of producing text within images and human aspects like hands.
In early October 2023, premium ChatGPT Plus users and Enterprise clients will be able to purchase DALL-E 3 as a research preview. Later this year, it will be made accessible to all users via the OpenAI labs website.
DALL-E and ChatGPT were different programs prior to this introduction. However, because DALL-E 3 has been developed from the ground up on ChatGPT. Users may utilize the AI chatbot to generate pictures from prompts ranging from brief sentences to in-depth paragraphs.
This news follows Google’s recent strengthening of its chatbot Bard. With tighter integration into several of its most well-known consumer services and products. Including Gmail, Docs, YouTube, and Maps. Additionally, Google enhanced Bard’s visual search capabilities, enabling users to contribute photographs. Using Google Lens and receive Google Search images in answer to queries in any of the chatbot’s supported languages.
Avoiding False Information
Additionally, according to OpenAI, precautions have been taken to prevent DALL-E 3 from producing violent, sexual, or hostile content. For instance, the service will reject requests for public figure portrayals. It is also looking at how to make it easier for users to tell when an image was produced using AI.
Earlier this week, OpenAI unveiled the OpenAI Red Teaming Network. A project through which it is looking for professionals in many industries to review its AI models and support its risk assessment and mitigation efforts. At different phases of the model and product development lifecycle. The business stated that DALL-E 3 is also intended to refuse requests for photos in the fashion of contemporary artists. And that creators have the choice to opt out of having their images used to train future image-generating models at OpenAI.
Be First to Comment