Live data from Hacker News

DALL·E now available in beta

openai.com

101–110 of 579 posts

Re: DALL·E now available in beta

#101
post #50
post #34

Earlier quoted context omitted.

Makes me think of Ouroboros https://en.m.wikipedia.org/wiki/Ouroboros

Reminds me of https://en.wikipedia.org/wiki/Low-background_steel

In this situation, the low-background steel is the MS-COCO dataset, associated with the Fréchet inception distance computed by comparing the statistical divergence between the high-level vector outputs of passing MS-COCO images through Google’s InceptionV3 classifier, and passing DALL-E images (or its competitors) through it.

For now at least, there is a detectable difference in variety.

Re: DALL·E now available in beta

#102
post #32

I find it amusing that they suggest DALL-E, which typically generates lovecraftian nightmare images, for making children's story illustrations.

yeah. dalle is "so bad it's good". it's great for post-post-ironic memes, but I don't see it being useful for anything else

No wireless. Less space than a nomad. Lame.

Re: DALL·E now available in beta

#103
post #31

Earlier quoted context omitted.

I don't see why there's any credible reason to expect that DALL-E will do anything at all to help those promoting the NFT silliness. Two separate issues.

If OpenAI could make a profit selling Dall-E images as NFT, I'd assume they'd do it, yeah?

Altman tried his hand at that by launching Worldcoin, and it didn't go well at all.

So I think it's prudent that OpenAI keep the 'sell shovels' business model instead with DALLE and GPT, at least for the time being.

Re: DALL·E now available in beta

#105

I am thrilled about DALL-E, and the new terms of service. However, how they implemented the improved "diversity" is hilarious. Turns out that they randomly, silently modify your prompt text to append words like "black male" or "female". See https://twitter.com/jd_pressman/status/1549523790060605440 I don't know which emotion I feel more - applause at how glorious this hack is or tears at how ugly it is. Good luck to…

> Turns out that they randomly, silently modify your prompt text to append words like "black male" or "female". I wonder what the distribution of those modifications is?

In my little testing, diversity in ethnicities was achieved but not realistic given the context. I also got a few androgynous people as I asked for a male or a female and another gender was appended.

Re: DALL·E now available in beta

#106
post #78

Earlier quoted context omitted.

They literally just add "black" and "female" with some weight before any prompt containing person. A comical work around to so called "bias" (isn't the whole point of these models to encode some bias?). Here's some experimentation showing this. https://twitter.com/rzhang88/status/1549472829304741888 As competitors with lower price points prop up, you'll see everyone ditch models with "anti bias" measures and take the…

Add after . So you can see the added words by making a prompt like "a person holding a sign saying ", and then the sign says the extra words if they are added.

Yeah actually, good call. The position of the token matters, since these things use transformers to encode the embeddings.

https://www.assemblyai.com/blog/how-imagen-actually-works/

Re: DALL·E now available in beta

#107

Interesting. I got access couple weeks ago (was on waitlist since the initial announcement) and frankly as much as really want to be excited and like it, DALL-E ended up being a bit underwhelming. IMHO - often results that produced are of low quality (distorted images, or quite wacky representation of the query). Some styles of imagery are certainly a better fit for being generated by DALL-E, but as far as commercial…

I felt the same way. If anything, I realized how soulless and uninteresting faceless art is. Dall-E 2 goes out of its way to make terrible faces for, im guessing, deepfake reasons?

Re: DALL·E now available in beta

#108
post #84

I am thrilled about DALL-E, and the new terms of service. However, how they implemented the improved "diversity" is hilarious. Turns out that they randomly, silently modify your prompt text to append words like "black male" or "female". See https://twitter.com/jd_pressman/status/1549523790060605440 I don't know which emotion I feel more - applause at how glorious this hack is or tears at how ugly it is. Good luck to…

it's a hard problem. at least they tried.

Honestly I would rather that they not try. I don't understand why a computer tool has to be held to a political standard.

Re: DALL·E now available in beta

#109

I am thrilled about DALL-E, and the new terms of service. However, how they implemented the improved "diversity" is hilarious. Turns out that they randomly, silently modify your prompt text to append words like "black male" or "female". See https://twitter.com/jd_pressman/status/1549523790060605440 I don't know which emotion I feel more - applause at how glorious this hack is or tears at how ugly it is. Good luck to…

as far as I can tell, they also concatenate "On LSD" to every prompt as well.

Re: DALL·E now available in beta

#110
post #84

I am thrilled about DALL-E, and the new terms of service. However, how they implemented the improved "diversity" is hilarious. Turns out that they randomly, silently modify your prompt text to append words like "black male" or "female". See https://twitter.com/jd_pressman/status/1549523790060605440 I don't know which emotion I feel more - applause at how glorious this hack is or tears at how ugly it is. Good luck to…

it's a hard problem. at least they tried.

It's not a "problem," it's an unwanted shard of reality piercing through an ideological guise.
Post reply on HN