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

DALL·E 2 prompt book [pdf]

dallery.gallery

Post reply on HN
Print view View original

121–130 of 157 posts

Re: DALL·E 2 prompt book [pdf]

#121

Post by tkgally »

tasuki wrote: post #112
tkgally wrote: post #60

Like many others, it seems, I have also been blown away by DALL-E 2. When I got access on Sunday, I first tried a lot of different prompts and got some interesting results. One semirandom one, “A photograph of a professor playing a grand piano on a rainy night in Tokyo,” produced some very atmospheric images. I then went down a rabbit hole of variations on that prompt (“A painting of...,” “A line drawing of...,” “A p…

The images are very cool, but... uhh, why turn them into a video? You know there are ways to share images on the internet?

Thank you for your comment and question. Out of vanity, I wanted the images to be accompanied by my piano music, however inept that music may be.

Please feel free to take screenshots of the images in the video and share them yourself on the Internet—giving appropriate credit to OpenAI, of course. You don’t need to credit me.

You may also run the same prompts through DALL-E 2 yourself and post the resulting images online; those images will be different from what I got. Or you could come up with your own prompts and post those images. In either case, please post the link here. I will be very interested in seeing the images you get.

Re: DALL·E 2 prompt book [pdf]

#122

Post by nbzso »

If you are commercial digital-painting illustrator or 3d illustrator artist, this is the moment to invest your time in other field. It is over.

Some will say that this 'tool' will help you in your creative process. In my obviously "biased" view, in the next 2-3 years, this will lower your monetary reward in half and create more requirements for competition with AI. People already are comparing DALL-E with human results.

In the long run, the software industry will eat itself to oblivion. Greed has no boundaries, and optimization of costs for corporations will never stop.

Some of my art related colleagues saw this 'trend' early and pivoted to crafts with added value for customers in the real world. On the oil painting side (which I am at) I don't feel any form of pressure, I paint for myself as a therapy. So, good luck:)

Re: DALL·E 2 prompt book [pdf]

#123

Post by archontes »

Siira wrote: post #55
archontes wrote: post #40

Earlier quoted context omitted.

This is absolutely the future, but it's not going to be obscure. You won't have a job if it isn't this or physical. AIs are going to replace entry level creatives, and experienced users with taste will largely perform selection and the development of good starts to mature designs. And I mean all creatives. Engineers, architects, mathematicians, programmers.

Mathematicians, doubtful. Someone must do some verification of the AI results. Programmers, also doubtful. You still need to design some APIs and interact with them. Interacting with an AI using natural language might become possible, but it definitely won’t be as efficient as more structured languages. (E.g., writing an algorithm in actual code is often much easier than teaching it to a human.)

It's exactly that verification of results and development from snippets and starts into a complete whole that will be the purview of human work. The AI will be replacing (or, rather, preventing the creation of positions for) juniors.

Re: DALL·E 2 prompt book [pdf]

#124

Post by hiidrew »

Angostura wrote: post #49
ehsankia wrote: post #32

Earlier quoted context omitted.

I've been using Midjourney [1] (it's not free, 25 photos demo, then 10$ for ~200 photos or 30$ for unlimited I think?). It's fairly fast, ~20s for a grid of 4, then as much for upscaling. I like the controls, it lets you do variations and tweak the image as you go. It's not as good for doing concrete asks, but it's very good for getting specific vibes. The website feed [1] requires Discord login to view examples, but…

What's the business with Midjourney demanding Discord details to sign up? Seems off.

It is kind of annoying but lurking and seeing other people's pictures is nearly as enjoyable as making your own. Plus you can use the bot in your dms if you'd like to keep them private.

Re: DALL·E 2 prompt book [pdf]

#125

Post by hiidrew »

soperj wrote: post #17
jw1224 wrote: post #13

I’ve spent just over 1 week with DALL•E 2. Over the past 7 days I’ve generated ~1000 images, 150 of which were good enough to save. I only saved images which made me audibly gasp. Witnessing your own novel idea spring to life is a magical experience. DALL•E provides an artistic tool on a comparable level to digital photography, and by extension Photoshop. At this stage it’s 100% clear to me that DALL•E has heralded i…

Honestly, I have a great use case for it currently, but then I realized it can only do square pictures, when I really want something that is much wider than it is tall.

In midjourney you can specify the aspect ratio, I think there is an example in their docs

Re: DALL·E 2 prompt book [pdf]

#126

Post by atwood22 »

cheschire wrote: post #71
atwood22 wrote: post #63

Earlier quoted context omitted.

Most people are objective dumb (i.e. less smart than the median/average) so this doesn’t surprise me.

How can most of any quantity be less than average? Edit: ah, I think I was thrown off by the use of “median/average” as equivalents.

The median and average are equivalent in a normal distribution.

Re: DALL·E 2 prompt book [pdf]

#127

Post by tasuki »

tkgally wrote: post #121
tasuki wrote: post #112

Earlier quoted context omitted.

The images are very cool, but... uhh, why turn them into a video? You know there are ways to share images on the internet?

Thank you for your comment and question. Out of vanity, I wanted the images to be accompanied by my piano music, however inept that music may be. Please feel free to take screenshots of the images in the video and share them yourself on the Internet—giving appropriate credit to OpenAI, of course. You don’t need to credit me. You may also run the same prompts through DALL-E 2 yourself and post the resulting images onl…

> Out of vanity, I wanted the images to be accompanied by my piano music, however inept that music may be.

I liked the music! Didn't know it was yours :)

I'd still prefer to see the images in a way that enables me to choose how long to look at each. It could be a webpage that plays your music in the background?

> Please feel free to take screenshots [...] You may also run [...]

Well, I wanted a low effort way to view the images, but thank you :)

Re: DALL·E 2 prompt book [pdf]

#128

Post by cheschire »

atwood22 wrote: post #126
cheschire wrote: post #71

Earlier quoted context omitted.

How can most of any quantity be less than average? Edit: ah, I think I was thrown off by the use of “median/average” as equivalents.

The median and average are equivalent in a normal distribution.

But then would you still be able to say “most” are below average of the distribution is normal?

Re: DALL·E 2 prompt book [pdf]

#129

Post by autoexec »

ehsankia wrote: post #37

Prompt crafting is quickly becoming an art. I just found out yesterday that there's actually market places for buying and selling prompts [0]. It can really make a big difference if you can tune the image by adding the right words. Midjourney [1] even allows things such as adjusting the weight of each keyword or how "literal" the AI should take your prompt. [0] https://promptbase.com/ [1] https://midjourney.gitbook.i…

That's kind of depressing really. It's like paying someone for google search terms. I remember when the internet was full of sites filled with interesting and effective google search terms (called "dorks") but nobody was charging $2 for each one, they were just sharing something cool with the world and helping everyone use the tool more effectively.

Re: DALL·E 2 prompt book [pdf]

#130

Post by mojojojo24 »

meowtastic wrote: post #101
mojojojo24 wrote: post #81

Earlier quoted context omitted.

If you're using Midjourney, you might like this prompt generator that I made after scraping prompts from their Discord server: https://huggingface.co/succinctly/text2image-prompt-generato...

Looks cool! One piece of feedback I have is that I personally wouldn't want to autocomplete prompts as shown in the image you have there. I'd want something that "fixes" my prompts so that they're understood by the model. Like say if I put "Afro-Asian samurai carrying two swords in a cyberpunk environment", for it to say how I could re-word it to generate something closer to what I actually want.

That's a really neat idea, since most of the time people already know what they want to generate. I'd have to think a little more about how I'd get the training data for that. LMK if you have any ideas.
Post reply on HN