DALL-E Now Available Without Waitlist
OpenAI (tweet, via Hacker News):
Starting today, we are removing the waitlist for the DALL·E beta so users can sign up and start using it immediately. More than 1.5M users are now actively creating over 2M images a day with DALL·E—from artists and creative directors to authors and architects—with over 100K users sharing their creations and feedback in our Discord community.
[…]
We are currently testing a DALL·E API with several customers and are excited to soon offer it more broadly to developers and businesses so they can build apps on this powerful system.
It’s really amazing how DALL-E missed the boat. When it was launched, it was a truly amazing service that had no equal. In the months since then, both Midjourney and Stable Diffusion emerged and got to the point where they produce images of equal or better quality than DALL-E. And you didn’t have to wait in a long waitlist in order to gain access! They effectively gave these tools free exposure by not allowing people to use DALL-E.
Furthermore, the pricing model is much worse for DALL-E than any of its competitors. DALL-E makes you think about how much money you’re losing continuously - a truly awful choice for a creative tool! Imagine if you had to pay photoshop a cent every time you made a brushstroke. Midjourney has a much better scheme (and unlimited at only 30/month!), and, of course, Stable Diffusion is free.
I’ve usually heard that DALL-E’s quality is much better.
Previously:
2 Comments RSS · Twitter
Yes, Dalle seems to be under the impression that Francis Bacon is a photorealist. I'm pretty fed up with the Midjjourney look too, but at least it doesn't give me nightmares.
"I’ve usually heard that DALL-E’s quality is much better."
It's not. I haven't used Midjourney much, because the interface is annoying, but between Stable Diffusion to DALL-E, there's really no comparison. DALL-E tends to produce insane LSD-fueled wackiness most of the time, with very obvious AI artifacts. You have to generate a dozen images to get two usable ones. Stable Diffusion, on the other hand, almost consistently produces output that looks like a photograph, or something produced by an actual person.
A friend of mine made the comparison that Stable Diffusion knows what humans want and tries to please them, while DALL-E seems to just be doing its own thing for its own amusement.
I think Stable Diffusion might be a little bit worse at interpreting the intent of the prompt than DALL-E, but that's kind of hard to tell, because both of them are likely to get confused about similar things.