Logged-out Icon

Home / Consumer Tech

OpenAI introduces DALL-E 3, an upgraded text-to-image tool integrating ChatGPT

DALL-E 3 from OpenAI not only creates better-quality images but also prioritizes artist rights and tackles algorithmic biases

DALL-E 3

OpenAI has just introduced DALL-E 3, a revamped version of its renowned text-to-image tool, DALL-E. This new version integrates features from ChatGPT, OpenAI’s trending AI chatbot, to make the image prompting process smoother.

Today’s top-notch AI-driven image generators can create visual art based on textual prompts. These prompts describe the image’s theme and can be anything from photorealistic to whimsical designs. But, there’s a catch. Crafting the perfect prompt is more art than science, to the point where “prompt engineering” is emerging as a legitimate profession.

Enter DALL-E 3. With an integration of ChatGPT, OpenAI has added a conversational touch. Subscribers of OpenAI’s ChatGPT Plus and ChatGPT Enterprise services can just type in what kind of image they want. The chatbot assists in refining the request, making the prompt more precise and descriptive. What’s impressive is that even a brief prompt can be expanded upon by ChatGPT, ensuring that DALL-E 3 has a clear idea of what to generate.

But DALL-E 3 isn’t just about ChatGPT. OpenAI claims that this version creates better-quality images that stay true to the given prompts, especially the lengthy ones. Plus, it’s getting smarter. Earlier image generation tools often struggled with intricate details, like human hands or textual content. DALL-E 3 seems to have a firmer grip on these challenges.

And OpenAI didn’t stop at quality. Addressing rising concerns about algorithmic biases and digital safety, DALL-E 3 comes with advanced features. For instance, it won’t generate images resembling the style of living artists or featuring public figures. A significant step toward respecting creators’ rights is the option for artists to opt out, preventing their work from being used in training future OpenAI text-to-image tools. This move might be in response to recent controversies where AI companies, including OpenAI, faced accusations of using copyrighted art to train their AI models.

The release of DALL-E 3 is timely. The world of generative AI is bustling with innovation, especially when it comes to synthesizing images. Companies like Midjourney and Stability AI are stepping up their game, compelling OpenAI to ensure it remains a front-runner.

As for availability, OpenAI will first offer DALL-E 3 to its premium ChatGPT users come October. Research institutions and API clientele are next in line. What remains uncertain, however, is if OpenAI intends to provide a complimentary online tool for DALL-E 3, akin to what was available for DALL-E 2 and its predecessor.

Advertisements