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

361–370 of 579 posts

Re: DALL·E now available in beta

#361

Post by smusamashah »

speedgoose wrote: post #334
smusamashah wrote: post #286

Earlier quoted context omitted.

They won't. DALL-E images are mostly not as high quality. The high quality stuff which everyone has been sharing is result of lots of cherry picking.

In my experience it doesn’t require that much cherry picking if you use a carefully crafted prompt. For example: “ A professional photography of a software developer talking to a plastic duck on his desk, bright smooth lighting, f2.2, bokeh, Leica, corporate stock picture, highly detailed” And this is the first picture I got: https://labs.openai.com/s/lSWOnxbHBYQAtli9CYlZGqcZ It got it a bit strong on the depth of fi…

Even this image is just an illusion of a perfect photo, which is a blur for most part, see the face of duck. I had access since past 4 5 days and it fails badly whenever I tried to create any unusual scene.

For the first few days when it was announced I use to look deep even in real photos in search of generative artifacts. They are not so difficult to spot now, most of the times anyway.

Re: DALL·E now available in beta

#362

Post by hit8run »

It’s so dirty what Microsoft is doing here. They ripped the tech out of developers hands just to sell us drips of it. Drips that are not enough to build a product for more than a few people. They require to check on the use before launching etc. I truly hate this company, their shitty operating system and their monopoly business game. Everything they buy turns to shit. And don’t tell me about VSCode. It’s just a trap to fool developers.

Re: DALL·E now available in beta

#363

Post by fishtoaster »

zionic wrote: post #343
isoprophlex wrote: post #296

Wait until someone trains a model like this, for porn. There seems to be a post-DALLE obscenity detector on openAI's tool, as so far I've found it to be entirely robust against deliberate typos designed to avoid simple 'bad word lists'. Ask it for a "pruple violon" and you get purple violins... you get the deal. "Metastable" prompts that may or may not generate obscene (content with nudity, guns, violence as I've fou…

Honestly that part pisses me off. Who cares if their AI "makes porn" or something "offensive".

I suspect it's more a business restriction than a moral one. If OpenAI allows people to make porn with these tools, people will make a ton of it. OpenAI will become known as "the company that makes the porn-generating AIs," not "the company that keeps pushing the boundaries of AI." Being known as the porn-ai company is bad for business, so they restrict it.

Re: DALL·E now available in beta

#364

Post by woojoo666 »

dplavery92 wrote: post #348
TigeriusKirk wrote: post #252

Earlier quoted context omitted.

What are the resources at work here? What are the resources needed to train this model? If someone just gave you the model for free, what resources would you need to use it to generate new results?

In the unCLIP/DALL-E 2 paper[0], they train the encoder/decoder with 650M/250M images respectively. The decoder alone has 3.5B parameters, and the combined priors with the encoder/decoder are the in the neighborhood of ~6B parameters. This is large, but small compared to the name-brand "large language models" (GPT3 et. al.) This means the parameters of the trained model fit in something like 7GB (decoder only, half-p…

> This means the parameters of the trained model fit in something like 7GB (decoder only, half-precision floats) to 24GB (full model, full-precision)

> you would probably want an enterprise cloud/data-center GPU like an NVIDIA A100, especially if running batches of more than one image.

That doesn't seem so bad.

looks up price of NVIDIA A100 - $20,000

oh...ok I'll probably just pay for the service then

Re: DALL·E now available in beta

#365

Post by Pxtl »

mikeyouse wrote: post #49
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.

This precise thing is causing a funny problem in specialty areas. People are using e.g. Google Lens to identify plants, birds and insects, which sometimes returns wrong answers e.g. say it sees a picture of a Summer Tanager and calls it a Cardinal. If the people then post "Saw this Cardinal" and the model picks up that picture/post and incorporates it into its training set, it's just reinforcing the wrong identificat…

Then that's a cardinal now.

Re: DALL·E now available in beta

#366

Post by peteforde »

nsxwolf wrote: post #234
peteforde wrote: post #228

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…

I'm already bored of it. When you have everything, you have nothing.

I don't know how to say this without sounding like a jerk, even if I bend over backwards to preface that this isn't my intent: this statement says more about your creativity and curiosity than a ceiling on how entertaining DALL-E can be to someone who could keep multiple instances busy, like grandma playing nine bingo cards at once.

Knowing that it will only get better - animation cannot be far behind - makes me feel genuinely excited to be alive.

Re: DALL·E now available in beta

#367

Post by pkaye »

jnovek wrote: post #316
peteforde wrote: post #228

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…

My heart sank when I saw the pricing model. I’ve been creating generative art since 2016 and I’ve been anxiously waiting for my invite. I wont be able to afford to generate the volume of images it takes to get good ones at this price point. I can afford $20/mo for something like this but I just can’t swing $200 to $300 it realistically takes to get interesting art out of these CLIP-centric models. Heck, the initial 5…

I'm sure the prices will go down each year as the computing costs go down.

Re: DALL·E now available in beta

#368

Post by arrow7000 »

ajafari1 wrote: post #134

I wrote about this happening two days ago on my sub stack post, "OpenAI will start charging businesses for images based on how many images they request. Just like Amazon Web Services charges businesses for usage across storage, computing, etc. Imagine a simple webpage where OpenAI will list out their AI-job suite, including “jobs” such as software developer, graphics designer, customer support rep, and accountant. Yo…

"Business monetises their offering" can't say I'm entirely blown away by the prediction

Re: DALL·E now available in beta

#369

Post by uxamanda »

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.

The part where you have to confirm you are not a robot to subscribe to the mailing list is the best part of this, my new favorite website.

Re: DALL·E now available in beta

#370

Post by peteforde »

dplavery92 wrote: post #348
TigeriusKirk wrote: post #252

Earlier quoted context omitted.

What are the resources at work here? What are the resources needed to train this model? If someone just gave you the model for free, what resources would you need to use it to generate new results?

In the unCLIP/DALL-E 2 paper[0], they train the encoder/decoder with 650M/250M images respectively. The decoder alone has 3.5B parameters, and the combined priors with the encoder/decoder are the in the neighborhood of ~6B parameters. This is large, but small compared to the name-brand "large language models" (GPT3 et. al.) This means the parameters of the trained model fit in something like 7GB (decoder only, half-p…

Thanks for the really excellent insight and links.

I do hope that the conversation starts to acknowledge the difference between sunk costs and running costs.

Employees, office leases and equiment are all happening, regardless and ongoing.

Training DALL-E 2: very expensive, but done now. A sunk cost where every dollar coming in makes the whole endeavor more profitable.

Operating the trained model: still expensive, but you can chart out exactly how expensive by factoring in hardware and electricity.

I believe that by not explicitly separating these different columns when discussing expense vs profit, we're making it harder than it needs to be to reason about what it actually costs every time someone clicks Generate.

Post reply on HN