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

No elephants: Breakthroughs in image generation

oneusefulthing.org

Post reply on HN
Print view View original

151–160 of 373 posts

Re: No elephants: Breakthroughs in image generation

#151

Post by butlike »

masswerk wrote: post #128
wavemode wrote: post #103

Earlier quoted context omitted.

Companies like Studio Ghibli are not being harmed by AI, small freelance artists are.

I think, Studio Ghibli will be affected, as well, since their "trademark style" (as we used to say), formerly a welcome sight and indicative for a certain type of story telling, will be devaluated as an indicator for slop. (Much like there are certain traits of an image, which we associate with soap operas and assume to be indicative of a low-value production.)

I doubt that. "Which movie does the slop belong to?" "Oh none of them? Ok" Is a pretty easy search term

Re: No elephants: Breakthroughs in image generation

#152

Post by butlike »

SirMaster wrote: post #148
eadmund wrote: post #125

Earlier quoted context omitted.

It took a truly colossal amount of human time and effort to build AI systems. It takes significant amount of energy to run those AI systems. I don’t see any meaningful difference at all between the system of a human, a computer and a corpus of images producing new images, and the system of a human, a paintbrush, an easel, a canvas and a corpus of images producing new images. Emphasis on the new — copying is still cop…

You don't see a difference between a person spending years learning techniques to create art by hand, and spending months or years studying and practicing some famous artists style, and then spending days manually crafting drawing a single piece of artwork in the style and quality of the originals. The difference between that, and a person just entering a prompt to create some drawing in some style. The model looked…

The rules do change, but as a meritocracy as society simply decides to move on or not. There will be no cabal of artists who define how the rules will change. It will be organic. Like moving on from cave paintings to impressionism.

Re: No elephants: Breakthroughs in image generation

#153

Post by zwnow »

UncleEntity wrote: post #131
zwnow wrote: post #110

Earlier quoted context omitted.

People love to make things seem harder than they are. I tattoo people, I am aware about skin types, usually thats not a big issue unless its heavily scarred. The quality of your tattoo machine matters most, as my 70€ eBay makeshift one wasnt nearly as good as a proper one. Amount of ink matters, needle depth, skin type, sweat. But thats stuff you have figured put after your 20th tattoo. Its like knowing datatypes in…

> But thats stuff you have figured put after your 20th tattoo. So... after spending hundreds, if not thousands, of hours learning a skill? I got a tattoo back in the day and specifically went to one the guys in my platoon said was good due to him being featured in magazines or whatever. It's kind of an important thing to get right on the first, not 20th, attempt IMHO.

20 tattoos hundreds of hours? Lol.

Re: No elephants: Breakthroughs in image generation

#154

Post by zwnow »

bongodongobob wrote: post #134
flkiwi wrote: post #99

Earlier quoted context omitted.

A friend who is heavily inked has gone on at length to me about understanding skin elasticity--particularly how it changes over a lifetime--as well as the way joints and muscles change and distort visual lines, etc. It sure seems like a skilled trade to me. And, I don't know, depth of penetration of a needle in flesh and sanitation don't strike me as minor things to get right.

Penetration is practice, sanitation is basically use gloves and an autoclave.

Also swap needles and use some cream, people really overestimate how hard tattooing is

Re: No elephants: Breakthroughs in image generation

#155

Post by lupusreal »

The image annotated to explain why no elephants are possible is very amusing.

To me, this kind of image generation isn't very interesting for creating final products, but is extremely useful for communicating design intent to other people when collaborating on large creative projects. Previously I used crude "ms paint" sketches for this, which was much more tedious and less effective.

Re: No elephants: Breakthroughs in image generation

#156

Post by ziofill »

I usually agree with most of Gary Marcus' points, but I'd really like to hear his take on this. One of his examples is that "the system can't generate a horse riding an astronaut" and in fact I tried a lot in the past but it would always draw the astronaut on top of the horse. Well, here is the result now: https://postimg.cc/QFtRjbHM

Re: No elephants: Breakthroughs in image generation

#157

Post by flkiwi »

nasmorn wrote: post #149
flkiwi wrote: post #104

Earlier quoted context omitted.

That is "artists should be grateful to work for exposure" on a grand scale.

Except they didn’t do any work for the exposure. If a marketing agency had come up and executed the Ghiblify everything model as a PR stunt we would call it the most genius creative campaign of the decade

They did though. The studio engaged in tremendous amounts of work and created good will, in addition to their specific creative works. Their visual style is tied up in that good will. Use of the visual style for profit without consent is, at least ethically, misappropriation of another's value. And "You should be pleased I used your creative work because now more people will know about you and you will make a lot of money from this!" is one of the oldest defenses to misappropriation of creativity.

I'm not even mad. We do a terrible job in our society of valuing artists and creative people generally and in explaining the value of intangible things, especially something like good will. People have been misappropriating fonts and clipart and screenshots in presentations and posters and whatnot, duplicating clever branding ideas and the creative efforts of others, and so on for _decades_ if not longer, all without ill intent. It's something we need to fix and never will. But when that becomes a channel for another to directly profit, it begins to venture out of harmlessness.

Re: No elephants: Breakthroughs in image generation

#158

Post by wincy »

munksbeer wrote: post #150
Gasp0de wrote: post #121

Earlier quoted context omitted.

I love using LLMs to generate pictures. I'd call myself rather creative, but absolutely useless in any artistic craft. Now, I can just describe any image I can imagine and get 90% accurate results, which is good enough for the presentations I hold, online pet projects (created a squirrel-themed online math-learning game for which I previously would have needed a designer to create squirrel highschool themed imagery)…

>I love using LLMs to generate pictures. I'd call myself rather creative, but absolutely useless in any artistic craft. Now, I can just describe any image I can imagine and get 90% accurate results May I ask what you use? I'm not yet even a paid subscriber to any of the models, because my company offer a corporate internal subscription chatbot and code integration that works well enough for what I've been doing so fa…

I was generating pictures to use for a little game I made with my six and ten year old kids. They were so excited to see us go from idea to execution so quickly, they were laughing and we had a ton of fun. The only thing that disappointed me was I got throttled. We’d need to pay for API image gen to get it even faster.

I made a logo for an internal product that wouldn’t have had a logo otherwise at our company. I also make a lot of shitpost memes to my friends to trash talk in the long running turn based war game we’ve all been playing, like “make a cartoony image of a dog man and a Greek giant beating up a devil” and the picture it gave was just hilarious and perfect, like an old timey Popeye cartoon.

Two years ago I was spending three hours using local models like Stable Diffusion to get exactly what I wanted. I had to inpaint and generate 100 variations which would have been insanely expensive if I wasn’t powering it with my own hardware.

Now I get something good in minutes, it’s crazy really.

Re: No elephants: Breakthroughs in image generation

#159

Post by jsheard »

ziofill wrote: post #156

I usually agree with most of Gary Marcus' points, but I'd really like to hear his take on this. One of his examples is that "the system can't generate a horse riding an astronaut" and in fact I tried a lot in the past but it would always draw the astronaut on top of the horse. Well, here is the result now: https://postimg.cc/QFtRjbHM

Whenever one of these well known gotcha prompts gets "solved" the question is always whether they actually solved the underlying reason it used to fail, or did they just have a bunch of third-world workers tag pictures of horses and astronauts until the model started handling that specific example more reliably. As the saying goes, every measure which becomes a target becomes a bad measure.

Re: No elephants: Breakthroughs in image generation

#160

Post by renewiltord »

nowittyusername wrote: post #141

There is circumstantial evidence out there that 4o image manipulation isn't done within the 4o image generator in one shot but is a workflow done by an agentic system. Meaning this, user inputs prompt "create an image with no elephants in the room" > prompt goes to an llm which preprocesses the human prompt > outputs a a prompt that it knows works withing this image generator well > create an image of a room > and th…

The prompt enrichment thing is pretty standard. Everyone does that bit, though some make it user-visible. On Grok it used to populate to the frontend via the download name on the image. The image editing is interesting.
Post reply on HN