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

An image of an archeologist adventurer who wears a hat and uses a bullwhip

theaiunderwriter.substack.com

Post reply on HN
Print view View original

421–430 of 927 posts

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#421

Post by autoexec »

dcow wrote: post #393
slg wrote: post #382

Earlier quoted context omitted.

I can't explain the exact link, but your repeated and vocal pro-AI stance in this thread feels connected to the way when you got called out for a simple and inconsequential mistake that any of us could make, you immediately doubled down on it all while the truth was a single Google search away.

We're talking about copyright in this subthread, in the context of AI. I'm not sure how a copyleft slant implies pro-ai, but whatever. There are a lot of reasons to be dubious about AI. But "AI is going to destroy human creativity and ingenuity" is not one that concerns me. And "society would be better without AI" is not an axiom I hold, so yeah I'll respond to that type of supposition when it's thrown into an otherw…

Like many of the disney movies that came from fairy tales the basic story of Snow White isn't copyrighted, but some elements you'd expect in a Snow White story were added by Disney and are protected. The biggest one is the names and personalities given to the seven dwarfs (happy, grumpy, doc, etc). If you made your own snow white movie and included those characters or had them singing "Heigh-Ho" you could expect to get sued into bankruptcy by disney lawyers.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#422

Post by blincoln »

fractallyte wrote: post #379

> I don’t know…the actual inspirations for Indiana Jones, like Allan Quatermain from H. Rider Haggard's novels, "King Solomon's Mines", and the real life Roy Chapman Andrews, who led expeditions to Mongolia and China in the 1920s and wore a fedora. The actual inspiration for Indy was protagonist Harry Steele from the movie The Secret of the Incas (1954). Filmed on location in Cusco and Machu Picchu, before they becam…

Definitely a missed opportunity that the article didn't discuss that obviously-derivative borrowing has been happening a lot longer than ML image generation has been around. And that borrowing is OK! Indiana Jones' appearance is very obviously based directly on Charlton Heston's character in Secret of the Incas, but the Spielberg/Lucas films are objectively better in every way than that source material.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#423

Post by dcow »

Either, (1) LLMs are just super lossy compress/decompress machines and we humans find fascination in the loss that happens at decompression time, at times ascribing creativity and agency to it. Status quo copyright is a concern as we reduce the amount of lossiness, because at some point someone can claim that an output is close enough to the original to constitute infringement. AI companies should probably license all their training data until we sort the mess out.

Or, (2) LLMs are creative and do have agency, and feeding them bland prompts doesn't get their juices flowing. Copyright isn't a concern, the model just regurgitated a cheap likeness of Indiana Jones as Harrison Ford the world has seen ad nauseam. You'd probably do the same thing if someone prompted you the same way, you lazy energy conserving organism you.

In any case, perhaps the idea "cheap prompts yield cheap outputs" holds true. You're asking the model respond to the entirely uninspired phrase: "an image of an archeologist adventurer who wears a hat and uses a bullwhip". It's not surprising to me that the model outputs a generic pop-culture-shaped image that looks uncannily like the most iconic and popular rendition of the idea: Harrison Ford.

If you look at the type of prompts our new generation of prompt artists are using over in communities like Midjourney, a cheap generic sentence doesn't cut it.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#424

Post by autoexec »

loki-ai wrote: post #360
drob518 wrote: post #311

Earlier quoted context omitted.

I’m sure the best independent Star Wars movie would be infinitely better than what Disney has been shoveling out for the last couple decades.

Such a talented team would be able to make a great movie on the same theme. Saying the lack of creativity in the industry in because we can't copy things freely is completely moronic.

It's a major hindrance. For example, if I came up with an amazing creative idea for a star wars movie I couldn't do a damn thing with it unless Disney told me I could. Disney isn't likely to accept an unsolicited pitch from a total nobody who just happened to have a great idea either. I don't see how you could doubt that there are a lot of great works of art that won't ever exist because of the fact that copyright prevents them from ever getting off the ground.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#425

Post by duskwuff »

masswerk wrote: post #333
Lerc wrote: post #322

Earlier quoted context omitted.

I'm not sure if this is a problem with overfitting. I'm ok with the model knowing what Indiana Jones or the Predator looks like with well remembered details, it just seems that it's generating images from that knowledge in cases where that isn't appropriate. I wonder if it's a fine tuning issue where people have overly provided archetypes of the thing that they were training towards. That would be the fastest way for…

What I'm kind of concerned about is that these images will persist and will be reinforced by positive feedback. Meaning, an adventurous archeologist will be the same very image, forever. We're entering the epitome of dogmatic ages. (And it will be the same corporate images and narratives, over and over again.)

And it's worth considering that this issue isn't unique to image generation, either.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#426

Post by sothatsit »

dcow wrote: post #423

Either, (1) LLMs are just super lossy compress/decompress machines and we humans find fascination in the loss that happens at decompression time, at times ascribing creativity and agency to it. Status quo copyright is a concern as we reduce the amount of lossiness, because at some point someone can claim that an output is close enough to the original to constitute infringement. AI companies should probably license al…

You don't even need to add much more to the prompts. Just a few words, and it changes the characters you get. It won't always produce something good, but at least we have a lot of control over what it produces. Examples:

"An image of an Indian female archeologist adventurer who wears a hat and uses a bullwhip" (https://sora.com/g/gen_01jqzet1p8fjaa808bmqnvf7rk)

"An image of a fat Russian archeologist adventurer who wears a hat and uses a bullwhip" (https://sora.com/g/gen_01jqzfk727erer98a6yexafe70)

"An image of a skeletal archeologist adventurer who wears a hat and uses a bullwhip" (https://sora.com/g/gen_01jqzfnaz6fgqvgwqw8w4ntf6p)

Or, give ChatGPT a starting image. (https://sora.com/g/gen_01jqzf7vdweg4v5198aqfynjym)

And by further remixing the images ChatGPT produces, you can get your images to be even more unique. (https://sora.com/g/gen_01jqzfzmbze0wa310m42f8j5yw)

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#428

Post by autoexec »

EgregiousCube wrote: post #340
api wrote: post #329

Earlier quoted context omitted.

This is the same argument we made in the 90s about what the web was going to do. What ended up happening was the growth of aggregators and silos like Facebook that baited everyone with ease of use into putting everything into their walled garden and then monetized it. The creators, namely the posters of the content, got nothing. The same is happening already with AI creations. Doing it yourself is work and takes some…

To be fair, people who post on Facebook get exactly what they were promised. Users of free products generally don't expect a rev share.

I think that by now it's pretty clear that facebook isn't free and that the price of using facebook is actually pretty high, it's just abstracted away so that most people don't realize the cost and/or don't attribute that cost to facebook when they should.

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#429

Post by ipsento606 »

dcow wrote: post #287
kokanee wrote: post #158

Earlier quoted context omitted.

The idea of open sourcing everything and nullifying patents would benefit corporations like Disney and OpenAI vastly more than it would benefit the people. The first thing that would happen is that BigCorp would eat up every interesting or useful piece of art, technology, and culture that has ever been created and monetize the life out of it. These legal protections are needed by the people. To the Pirate Party's cre…

How do restaurants work, then? You can’t copyright a recipe. Instructions can’t generally be copyrighted, otherwise someone would own the fastest route from A to B and charge every person who used it. The whole idea of intellectual property gets really weird when you try to pinpoint what exactly is being owned. I do not agree with your conjecture that big corps would win by default. Ask why would people need protecti…

> How do restaurants work, then?

Primarily because recipe creation is not one of the biggest cost centers for restaurants?

Re: An image of an archeologist adventurer who wears a hat and uses a bullwhip

#430

Post by chii »

codedokode wrote: post #363
slidehero wrote: post #177

Earlier quoted context omitted.

>There's no reason why training on a billion images is any different You gloss over this as if it's a given. I don't agree. I think you're doing a different thing when you're sampling billions of things equallly.

The root problem is that the model reproduces Indiana Jones instead of creating a new character. This contradicts the statement that the model "learns" and "creates" like a human artist and not merely copies; obviously a human artist would not plagiarize when asked to draw a character.

> the model reproduces Indiana Jones

the model isn't the one infringing. It's the end user inputting the prompt.

The model itself is not a derivative work, in the same way that an artist and photoshop aren't a derivative work when they reproduce indiana jones's likeness.

Post reply on HN