Back in the fall of last year the buzz in AI imaging, then only beginning to take hold of the popular imagination, was all about Open AI's DALL-E 2. How long ago that now seems. Since then, Midjourney has deservedly been been recognized by most as the premiere text-to-image AI generative app while DALL-E 2 has been largely forgotten. It's no surprise. Even at the height of its popularity its imagery seemed, to me at least, markedly inferior to that which I obtained from it's main rival, Stability AI's Stable Diffusion.
In spite of the above, it seems Open AI has not yet given up on DALL-E and, at least according to a YouTube video I recently viewed, is preparing to launch version 3. The original video seems to have been removed since then, or I'm simply unable to locate it, but I remember thinking when I viewed it that the "leaked" images shown were not much better than those I'd seen generated by version 2. At any rate, there is no official word as of this writing from Open AI on any plans to update DALL-E. I'll of course post here if I hear anything definitive.
No comments:
Post a Comment