Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

DALL·E now available in beta

openai.com

Post reply on HN
Print view View original

221–230 of 579 posts

Re: DALL·E now available in beta

#221

Post by stoicjumbotron »

berberous wrote: post #20
brycethornton wrote: post #3

I'm blown away by this: "Starting today, users get full usage rights to commercialize the images they create with DALL·E, including the right to reprint, sell, and merchandise. This includes images they generated during the research preview." I assumed this was going to be the sticking point for wider usage for a long time. They're now saying that you have full rights to sell Dall-E 2 creations?

I think they are reacting to competition. MidJourney is amazing, was easier to get into, gives you commercial rights, and frankly I found more fun to use and even better output in most instances.

Really hope I get an invite for MidJourney soon. Been on the waitlist since March :(

Re: DALL·E now available in beta

#222

Post by lagrange77 »

bemmu wrote: post #58

I was supposed to be making a video game, but got a bit sidetracked when DALL·E came out and made this website on the side: http://dailywrong.com/ (yes I should get SSL). It's like The Onion, but all the articles are made with GPT-3 and DALL·E. I start with an interesting DALL·E image, then describe it to GPT-3 and ask it for an Onion-like article on the topic. The results are surprisingly good.

http://dailywrong.com/new-course-teaches-guinea-pigs-househo...

lol

Re: DALL·E now available in beta

#223

Post by TaylorAlexander »

l33t2328 wrote: post #211
TaylorAlexander wrote: post #70

Earlier quoted context omitted.

The monkey selfie was not derived from millions of existing works, and that is the difference. If an artist has a well-known art style, and this algorithm was trained on it and can copy that style, would the artist have grounds to sue? I don't know.

If I write a song am I not deriving it from the existing works I’ve been exposed to?

Sure but if you just release a basic copy of a Taylor Swift song you will get sued to oblivion. So the law seems (IANAL) to care about how similar your work is to existing works. DALL-E does not seem capable of showing you the work that influenced a result, so users will have no idea if a result might be infringing. What this means to me is that with many users, some of the results would be legally infringing.

Re: DALL·E now available in beta

#224

Post by can16358p »

nharada wrote: post #28

Something I haven’t seen anyone talking about with these huge models: how do future models get trained when more content online is model generated to start with? Presumably you don’t wanna train a model on autogenerated images or text, but you can’t necessarily know which is which.

I think with the terms requiring explicitly telling which images/parts were generated, they could be filtered out and prevent a feedback loop of "generated in/generated out" images. I'm sure there will be some illegal/against terms of use cases there but the majority should represent fair use.

Re: DALL·E now available in beta

#225

Post by arecurrence »

thorum wrote: post #182
xnx wrote: post #76

I fully expect stock image sites to be swamped by DALL-E generated images that match popular terms (e.g. "business person shaking hands"). Generate the image for $0.15. Sell it for $1.00.

DALLE images are still only 1024 px wide. Which has its uses, but I don’t think the stock photo industry is in real danger until someone figures out a better AI superresolution system that can produce larger and more detailed images.

You can obtain any size by using the source image with the masking feature. Take the original and shift it then mask out part of the scene and re-run. Sort of like a patchwork quilt, it will build variations of the masked areas with each generation.

Once the API is released, this will be easier to do in a programmatic fashion.

Note: Depending on how many times you do this... I could see there being a continuity problem with the extremes of the image (eg: the far left has no knowledge of the far right). An alternative could be to scale the image down and mask the borders then later scale it back up to the desired resolution.

This scale and mask strategy also works well for images where part of the scene has been clipped that you want to include (EG: Part of a character's body outside the original image dimensions). Scale the image down, then mask the border region, and provide that to the generation step.

Re: DALL·E now available in beta

#226

Post by TaylorAlexander »

ZetaZero wrote: post #96
TaylorAlexander wrote: post #70

Earlier quoted context omitted.

The monkey selfie was not derived from millions of existing works, and that is the difference. If an artist has a well-known art style, and this algorithm was trained on it and can copy that style, would the artist have grounds to sue? I don't know.

> If an artist has a well-known art style, and this algorithm was trained on it and can copy that style... A lawyer could argue that the algorithm is producing a derivative work of the copyrighted input.

Right but if that work isn’t significantly changed from the source, it could be ruled as infringement. DALL-E cannot tell the users (afaik) if a result is close to any source material.

Re: DALL·E now available in beta

#227

Post by arecurrence »

redox99 wrote: post #177
xnx wrote: post #76

I fully expect stock image sites to be swamped by DALL-E generated images that match popular terms (e.g. "business person shaking hands"). Generate the image for $0.15. Sell it for $1.00.

DALL-E 2 isn't good enough for such photorealistic pictures with humans as of yet however.

There has been trouble with generating life-like eyes but a second pass with a model tuned around making realistic faces has been very successful at fixing that.

Re: DALL·E now available in beta

#228

Post by peteforde »

I have been having a blast with DALL-E, spending about an hour a day trying out wild combinations and cracking my friends up. I cannot imagine getting bored of it; it's like getting bored with visual stimulus, or art in general.

In fact, I've been glad to have a 50/day limit, because it helps me contain my hyperfocus instincts.

The information about new pricing is, to me as someone just enjoying making crazy imagines, a huge drag. It means that to do the same 50/day I'd be spending $300/month.

OpenAI: introduce a $20/month non-commercial plan for 50/day, and I'll be at the front of the line.

Re: DALL·E now available in beta

#229

Post by bequanna »

naillo wrote: post #117
cypress66 wrote: post #35

> Reducing bias: We implemented a new technique so that DALL·E generates images of people that more accurately reflect the diversity of the world’s population. This technique is applied at the system level when DALL·E is given a prompt about an individual that does not specify race or gender, like “CEO.” Will it do it "more accurately" as they claim? As in, if 90% of CEOs are male, then the odds of a CEO being male i…

It's also funny that this likely won't 'unbias' any actual published images coming out of it. If 90% of the images in the world has a male CEO, then for whatever reason that's the image people will pick and choose from DALL-Es output. (Generalized to any unbiasing - i.e. they'll be debiased by humans.)

Imagine you're in South Korea (or any other ethnically homogenous country). Do you want "black" "female" randomly appended to your input?

Re: DALL·E now available in beta

#230

Post by throwaway4aday »

time_to_smile wrote: post #171
danielvf wrote: post #80

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…

This is funny because I work on a team that is using GPT-3 and to fix a variety of issues we have with incorrect output we've just been having the engineering team prepend/append text to modify the query. As we encounter more problems the team keeps tacking on more text to the query. This feels like a very hacky way to essentially reinvent programming badly. My bet is that in a few years or so only a small cohort of…

How else would you specify the type of image you would like? Surely, if you were hiring a designer you would provide them with a detailed description of what you wanted. More likely, you would spend a lot of time with them maybe even hours and who knows how many words. For design work specifically to create a first mockup or prototype of a product or image it seems like DALL-E beats that initial phase hands down. It's much easier to type in a description and then choose from a set of images than it is to go back and forth with someone who may take hours or days to create renderings of a few options. I don't think it'll put designers out of work but I do think they'll be using it regularly to boost their productivity.
Post reply on HN