OpenAI boasted third generation image AI generator DALL-E. At one time, it was a technological leader, but the company never turned it into a service as accessible and high-quality as the competing Midjourney.
Now that could change, as the DALL-E 3 will arrive during October as an exclusive feature for the ChatGPT Plus subscriberscorporate clients and, of course, also to platform for developers.
They say no more complex prompt engineering
What DALL-E 3 can do remains to be seen, but OpenAI promises it will much more creative and will pay attention to the wording of the entire text command.
DALL-E 3 promises to take care of all the information that is in the text command when generating. At the same time, prompting should be much more natural
The authors on the blog explicitly mention that the goal should be that we no longer have to deal with it prompt engineering – often quite complex and unnaturally constructed text commands that force the generator to really create what we need.
You can find the first samples both in the blog post and on Instagram or perhaps twitter account Sam Altman, head of OpenAI. And since he added prompts to some of the images, we tried entering the same prompt into the Midjourney service. We can compare the results:
Vlevo DALL-E 3, vpravo Midjourney. Zadání: An illustration of an avocado sitting in therapist’s chair, saying “I just feel so empty inside” with a pit-sized hole in its center. The therapist, a spoon, scribbles notes.
Vlevo DALL-E 3, vpravo Midjourney. Zadání: An illustration of a human heart made of translucent glass, standing on a pedestal amidst a stormy sea. Rays of sunlight pierce the clouds, illuminating the heart, revealing a tiny universe within. The quote ‘Find the universe within you’ is etched in bold letters..
If OpenAI chose real images for the DALL-E presentation, it is obvious that it can work well with text. Getting Midjourney to add some writing is often impossible.
2023-09-20 18:49:20
#OpenAI #introduced #DALLE #image #generator #arrive #ChatGPT #days
