OpenAI has introduced an upgraded version of its text-to-image tool, DALL-E 3, which leverages the capabilities of ChatGPT, the widely recognized AI chatbot, to simplify the process of generating prompts.
Modern AI-powered image generation tools transform textual prompts, describing images in various styles ranging from photorealistic to fantastical. However, crafting the perfect prompt can be a formidable task, to the extent that “prompt engineering” has emerged as a legitimate profession.
OpenAI’s latest offering, DALL-E 3, integrates ChatGPT to assist in refining prompts. Subscribers to OpenAI’s premium ChatGPT plans, ChatGPT Plus and ChatGPT Enterprise, can now input a request for an image and fine-tune it through interactions with the chatbot. The results are conveniently delivered within the chat application.
ChatGPT is adept at taking even brief prompts and enriching them with additional details, providing enhanced guidance to the DALL-E 3 model.
The integration with ChatGPT isn’t the sole enhancement in DALL-E 3. OpenAI claims that DALL-E 3 generates higher-quality images that closely align with the provided prompts, especially when dealing with more extensive prompts. It also exhibits improved handling of content that has historically challenged image-generating models, such as text and depictions of human hands.
In addition to these improvements, DALL-E 3 incorporates new mechanisms aimed at reducing algorithmic bias and enhancing safety. For instance, it will decline requests for images mimicking the styles of living artists or portraying public figures. Moreover, artists now have the option to opt out of having their artwork used to train future iterations of OpenAI’s text-to-image models. (OpenAI, like some of its competitors, has faced legal action over allegations of using copyrighted artwork to train its generative AI image models.)
The launch of DALL-E 3 coincides with the intensifying competition in the generative AI sector, particularly in the realm of image synthesis. Competitors such as Midjourney and Stability AI continue to refine their image-generating models, placing pressure on OpenAI to maintain its leadership position.
OpenAI’s plan is to initially introduce DALL-E 3 to premium ChatGPT users in October, followed by its availability to research labs and API customers. The company has not disclosed whether it intends to release a free web tool, similar to its approach with DALL-E 2 and the original DALL-E model.