post pictures your computer has generated ( algorithmic art )

C

catalog

Well-known member
What they've done is a limit on dissonance so people can make a useful image rather than anything troublesome
 
C

catalog

Well-known member
If you imagine the text prompt as a fairly shallow level of variance, cos ultimately its base on language, you can gget good results off not much if you also narrow the bandwidth
 
william_kent

william_kent

Well-known member
catalog said:
What they've done is a limit on dissonance so people can make a useful image rather than anything troublesome
Click to expand...

my opinion is that because openAI own the copyright to any images produced by DALL-E 2, they don't want lawsuits just because some people want to make their own niche special interest pr0n, or to generate images of famous, possibly litigious, people in compromising positions, "deep fakes", hateful images, images in the style of artists whose work is still in copyright who, or their estate, could claim infringement, or anything else that could possibly land the corporation in hot water

edit: TLDR; AI being neutered by lawyers
 
Last edited:
C

catalog

Well-known member
sufi

sufi

lala
william_kent said:
my opinion is that because openAI own the copyright to any images produced by DALL-E 2, they don't want lawsuits just because some people want to make their own niche special interest pr0n, or to generate images of famous, possibly litigious, people in compromising positions, "deep fakes", hateful images, images in the style of artists whose work is still in copyright who, or their estate, could claim infringement, or anything else that could possibly land the corporation in hot water

edit: TLDR; AI being neutered by lawyers
Click to expand...
i wonder what the data of all those queries is worth?
it's very fascinating to look at the odd shit people want the ai to draw, the requests it elicits - what does the sum of all these hooman fantasies represent??
 
C

catalog

Well-known member
Members of the 'Dissensus' internet forum argue about the relative merits of different strains of UK rave music, in the style of the painter George Shaw

DoBwkVH.png
 
william_kent

william_kent

Well-known member
the cynical part of me thinks that openAI is using the people with dall-e 2 access as unpaid testers, but that is just because I'm jealous

@droid mentioned earlier that Midjourney is now in open beta , so now is your chance if you want to try it out for free ( same cynical thinking applies regarding being used as unpaid testers ) - although I've ended up paying to get more GPU time because it's quite addictive
 
william_kent

william_kent

Well-known member
an announcement from openAI today:

DALL-E now available in beta

We’ll invite 1 million people from our waitlist over the coming weeks. Users can create with DALL·E using free credits that refill every month, and buy additional credits in 115-generation increments for $15.
 
william_kent

william_kent

Well-known member
Midjourney vs DALLE-2 - head to head!

I've shamelessly stolen @catalog's prompt that he used on DALLE-2 and used it on Midjourney

1658353765051.png

DJ Hype having a Baileys and whipped Cream in Gillet Square, in the style of Gustav Klimt

slightly better likeness, and more in the style of?
 
william_kent

william_kent

Well-known member
microsoft have announced NUWA-Infinity

NUWA-Infinity is a multimodal generative model that is designed to generate high-quality images and videos from given text, image or video input

according to the abstract:

Compared to DALL-E, Imagen and Parti, NUWA-Infinity can generate high-resolution images with arbitrary sizes and support long-duration video generation additionally

there don't seem to be any links to a demo, or any indication of whether it will ever be available to the general public
 
You must log in or register to reply here.
Top