As Big Tech continues to jam every half-baked piece of artificial intelligence into every other over-baked piece of software, OpenAI has come full circle. The AI company behind ChatGPT is placing the wildly popular chatbot into the next version of its image generation tool.
Axios reports that OpenAI’s next generation of DALL-E is getting a ChatGPT upgrade. According to the outlet, the company says that DALL-E 3 will feature a better understanding of prompts over DALL-E 2, especially in regard to longer directives from users. The addition of ChatGPT will reportedly help DALL-E hone in on what users want the software to do through conversational dialogue entries. The souped-up DALL-E 3 will be released to ChatGPT+ and enterprise accounts this October, but there is no set date for a free version as reported by The Verge.
The company is also apparently trying to reign in artificial intelligence—just a teensy bit—with this new release. Notably, DALL-E 3 will reject prompts in which users ask the AI to create a piece of art that is modeled after a living artist’s style while dead artists are fair game. At the same time, artists will also be able to opt out of having their own work used to train future iterations of the generative AI—at least that’s what OpenAI claims. These new limits come as legal pressure against copyrighting AI work mounts in U.S. courts while AI art generators also face lawsuits over their training data.
OpenAI did not immediately return Gizmodo’s request for comment on the new DALL-E update.
OpenAI opened DALL-E to the wider world with the release of DALL-E 2 in September 2022—prior to that, the image generator was severely bottlenecked with a limited waitlist. Microsoft, one of the first Big Tech investors in OpenAI, announced in March of this year that DALL-E would be added to the company’s Bing in a bid to continue breathing life into the suffocating search engine.