OpenAI in April last year launched the second generation of DALL-E “Vincennes map” model, the model with the strength of the industry has attracted widespread attention, but also once out of the circle.
According to the foreign media Decoder, OpenAI is preparing the next generation of DALL-E AI model, the company is currently conducting a series of Alpha tests, and some users have been exposed to the AI model early.
According to the media, an anonymous user on Discord recently shared his experience in testing DALL-E 3.
The anonymous user said that the May test version of DALL-E 3 was already able to generate images with a variety of aspect ratios, as well as support for longer prompts and the ability to generate “normal text.
The DALL-E 3 model is being tested by OpenAI, which is expected to push AI image generation to the next level.
The July beta is even more powerful, but according to the anonymous user, the July beta is likely to randomly generate a series of NSFW scenes, as well as copyrighted trademarks, and OpenAI hopes to prohibit this kind of “inappropriate content” from being generated at the model level, so DALL-E 3 will need some time to make adjustments. Therefore, DALL-E 3 will need some time to make relevant training adjustments.
Foreign media said, DALL-E 3 in the image quality and diversity have a breakthrough, is expected to promote AI image generation into a new stage.
It was reported earlier, Stable Diffusion XL has launched 1.0 official version, Stability AI company claimed that Stable Diffusion XL 1.0 is the company’s “most advanced” AI tool so far, and the actual effect is indeed remarkable, so the DALL-E model of OpenAI is not only the most advanced AI tool, but also the most advanced AI tool. Stability AI claims that Stable Diffusion XL 1.0 is the company’s “most advanced” AI tool to date, and the results are impressive, so OpenAI’s DALL-E model will be going head-to-head.