What can I say, this is nothing short of amazing.
The input string was: “Silver Ferrari Dino 246 gts in studio with dramatic lighting”
And 20 seconds later 4 alternatives were delivered.
How an AI can synthesize an image like this based on how it understands a ton of other images is just mind boggling. It obviously understands shape, lighting, transparency etc, all within an imaginary space. Nothing modelled, nothing textured, no light calculations… Go figure…
None the less, this will be an important creative tool in the future, great for getting original and unique inspiration.
I’ve been playing with it and it is really nice thing. Next leap is to be able to generate 3D models from text prompts
Front wheel drive Dino? Heresy!
[It appears the wheel is spinning but the vehicle is stationary]
There’s a TV programme here in the UK called Fake or Fortune, where people try to establish that a painting they own is actually an Old Master with all that implies for its valuation. I think I’ll see if Dall-e can knock up a Monet or two for me.