111–120 of 205 posts
Interesting, how with billions of nodes and supposed "intelligence", the network hasn't been able to deduce a simple concept of symmetry in human faces. All of the eyes and all of the lips in all of the pictures are asymmetrical, which easily gives AI generated images away.
Earlier quoted context omitted.
The same thing happens for Vincent van Gogh, and presumably others too.
A good measure for whether you're more of a celebrity or an artist is how much of your face a google-image trained AI thinks belongs in your work.
That this happens at all is evidence that the training data hasn't been curated, cleaned, or labeled well enough.
Earlier quoted context omitted.
Well what if the artist has a popular self portrait?
Frida Kahlo comes to mind.
Again, if the training data was labeled well enough, confusion about this sort of thing shouldn't happen.
Hate to say it but when i see stuff like this it only reminds me of what we could have achieved if this ingenuity had been applied in another domain. Can't help feeling that this accidentally harms creative types and risks swamping us with visual junk. The technical achievment is astounding but no-one would seriously claim that crafting an image via a short prompt is creative except in the most cursory way. I'm proba…
People said this about cameras. About digital cameras. About digital photo editing software. The next generation will normalize these tools and find incredible ways to be creative within their new cutting edge medium. The post-art world is here! Just think about how history books will remember this period! The styles that will be borne of necessity, of the need to break down art and find what makes it tick.
Also about desktop publishing.
Remember all the printers (ie. people working in the printing industry operating printing machines) that were put out of a job when you could just buy a (electronic) printer for your home computer and just print whatever you wanted yourself?
People were wringing their hands about that too back then... now we take it for granted that we can instantly print whatever we want whenever we want, without having to pay an expensive professional to do it for us (something most people couldn't afford).
Has it resulted in more junk being printed? Absolutely. But it also let people print all sorts of fantastic not to mention useful things that would almost never have seen the light of day without cheap and easy access to home printers.
The xerox copier was similarly revolutionary... as was the printing press itself, which put a lot of scribes out of business.
Photoshop put a lot of airbrush artists out of business, and who does copy and paste with physical glue and paper anymore?
As with photography, printers, copiers and photoshop, artists who embrace this technology will be able to use it to enhance their creativity and speed up their creative process.
There'll be a lot more competition, a lot more junk but also a lot more fantastic art that we can't even dream of yet.
Earlier quoted context omitted.
"It" doesn't understand anything. It's just a very advanced madlibs engine based on a database of a billion of already known images.
Some people would probably argue that humans are essentially the same.
Earlier quoted context omitted.
They're sorted by last name, it took me a second too! I just wish I knew more of them.
Sorted by last name but listed by first name means you can start typing the artist's name and get to the right place in the list...
Interesting, how with billions of nodes and supposed "intelligence", the network hasn't been able to deduce a simple concept of symmetry in human faces. All of the eyes and all of the lips in all of the pictures are asymmetrical, which easily gives AI generated images away.
And even so, too symmetrical faces will look just as un-human as a face that is too asymmetric. You need a face that is just the right amount of symmetrical in order for it to actually look good.
I think you make it sounds simpler than what it is.
It's also not a model that is trained to make as realistic people as possible, it's trained on a lot of different things, so obviously it won't excel at making realistic people. But one can easily imagine that some future models will be heavily trained on making realistic people rather than semi-realistic everything, like Stable Diffusion is trained to do today.