The amazing AI DALL-E is now in beta

OpenAI opens the doors to the amazing DALL-E AI even more broadly. The beta version of the system, which produces stunning images from a text description, will be available to 1 million queued users.

It is impossible to miss the images produced by DALL-E if you are hanging out on social networks. These AI-generated illustrations from a text description in natural language are particularly successful and delight all users who have access to them. And there will be more of them, as OpenAI has announced that the system is available in beta.

1 million additional users

In the coming weeks, the 1 million interested people waiting in the DALL-E queue will have access to the AI. This trial will also give them 50 free credits for the first month of use, plus an additional 15 points each month. One credit equals four photos generated by the system, or three photos if it is a variation or modification of the photo.

Fans can also purchase 115 points (460 max images) for $15, in addition to their free credits. The consortium also details the licenses to exploit images generated by DALL-E: Users will retain full rights for commercial use, including copyright and sale rights. In other words, we can find DALL-E illustrations in children’s books, in newsletters, in games, in comics…

Be careful, not everything will be possible. OpenAI wants to reduce the risk of abuse of DALL-E. Images uploaded to the AI ​​feed will not be able to present realistic faces, and images of personalities and politicians will not be authorized. Moderation filters will prevent violent, “adult” or political images from being created.

OpenAI has also limited exposure of DALL-E models to explicit content. In contrast, AI is designed in a way that reflects the diversity of the population. To learn more about DALL-E, our article on the topic makes a whole point.

Leave a Comment