Highlights –

  • The team released yet another version in late August called “Outpainting,” which uses AI to expand an image similarly by leveraging natural language descriptions to add new images outside the confines of an image.
  • The revised system was initially previewed to consumers in April and was based on the initial version of DALL- E, which improved resolution while decreasing image creation latency.

The research team at OpenAI LLC revealed that the waitlist for its potent Artificial Intelligence (AI)-powered natural language-to-image generator DALL-E has been eliminated. Users may now sign up and begin using it right away.

The OpenAI team said in a statement, “More than 1.5 million users are now actively creating over 2 million images a day with DALL-E — from artists and creative directors to authors and architects—with over 100,000 users sharing their creations and feedback in our Discord community.”

It cited safety concerns over misuse and the responsibility of growing a complicated system as the reasons for the gradual introduction of the beta to the public and the adoption of an “iterative approach to deployment.” The revised system was initially previewed to consumers in April and was based on the initial version of DALL- E, which improved resolution while decreasing image creation latency.

The publication of the DALL-E white paper led to a tremendous amount of enthusiasm from the community about the full potential of AI picture generation. It was able to take text descriptions and turn them into surreal visuals of real-world scenes. Consequently, artists could use it to produce a wide range of art by learning how to define best what they wanted and how to instruct the AI to give it.

With these improvements, DALL-E has already provided the capability to change the images within uploaded images. This ability is known as “Inpainting.” It enables users to leverage text-to-image capabilities to change a specific area of an image. The team released yet another version in late August called “Outpainting,” which uses AI to expand an image similarly by leveraging natural language descriptions to add new images outside the confines of an image.

During the second beta period, the OpenAI team learned from practical examples how its AI generator could be misused and had already included guardrails to prevent misuse. It would not, for instance, replicate faces, sexual or violent material –in recent months, the team improved the filters to deny such requests better. The team stated, “In the past months, we’ve made our filters more robust at rejecting attempts to generate sexual, violent, and other content that violates our content policy and built new detection and response techniques to stop misuse.”

These filters were implemented to the platform to prevent concerns about AI-generated art posing ethical dilemmas, such as the creation of celebrity “deepfakes.” For example, though DALL-E can display beautiful landscapes and pictures of flocks of doves floating through the sunlight, it won’t produce a photorealistic representation of a famous person.

In this regard, other AI art generators, such as Midjourney and Stable Diffusion, are not the same. They don’t have the same filters, but they do block some content, and they can produce gruesome images like faces and erotica. These potentials have recently brought the morality of AI-generated art to the public’s attention.

DALL-E currently lacks an application user interface for programmers to use in their apps that are not part of the DALL-E platform, but the business stated in the blog post that it is testing one.

Even though the queue has been eliminated, DALL-E will still be in beta, and beta users’ prices won’t change. For their first month on the network, all users will receive 50 free credits, followed by 15 credits each subsequent month, with the opportunity to purchase more credits. Registrations are open right now.