🖼️ DALL-E 3 + ChatGPT im Test 👉🏻 Bilder in ChatGPT erstellen

The first paragraph introduces the viewer to the capabilities of D i3, a new AI image generation technology by Open AI, the company behind Jet GPT. The speaker discusses the ability to create photorealistic images of a character named Paula in a place called Lilaland with purple hair. It's highlighted that D i3 is a significant improvement over its predecessor, Dali 2, and is celebrated as an alternative to mid-journey. The paragraph also mentions that D i3 is available in the paid Plus version of chat GPT, not the free version, and demonstrates the ability to generate images using natural language prompts.

The second paragraph delves into testing D i3's ability to create consistent characters. The speaker requests the AI to generate an image of the previously discussed character, Paula, playing with a yellow ball. It's noted that the resulting image doesn't closely resemble the original character, prompting the speaker to ask for an image of the same person again. The paragraph also touches on the possibility of downloading the generated images and briefly mentions other AI image generation technologies like stable Diffusion XL.

The third paragraph explores D i3's capabilities further by testing it with different prompts, including creating an image of an American Muscle Car with a specific cinematic description. The speaker also attempts to generate an image of a cat drinking coffee, noting that while the results are not perfect, they are still quite good. The paragraph discusses the pros and cons of D i3, mentioning the high image quality and the ability to create consistent characters as advantages. However, it also points out the strict censorship as a significant drawback, where certain prompts are not allowed, even if they do not violate any explicit content policies.

The fourth paragraph focuses on the limitations of D i3, including the inability to use it in the free version of chat GPT and the presence of a digital watermark in the generated images, which cannot be removed even with post-processing. The speaker also mentions the strict content policies that prevent the creation of certain types of images. As an alternative, the paragraph suggests using Supermachine with stable Diffusion XL, which offers more freedom and flexibility without the same level of censorship. The speaker concludes by inviting viewers to share their thoughts on the Plus version of chat GPT and the new D i3 functionality.