Hacker News Reader

Comment by sandworm101 | Hacker News Reader

>> If the image is freely viewable (say you can browse to it), and you just look at it, are you violating any rights?

If I read Harry Potter, then turn around a write a book about a wizard with a z-shaped scar? Who works at a school for wizards? With a pet owl? Who is an orphan? At some point I have started to violate intellectual property rules. (Ignoring all the Harry Potter material that was itself lifted from prior public domain art.)

AI systems aren't just reading, they are generating material based on the stuff they have read. They and the people controlling them have to abide the copyright rules just like any other "author".

View on HN

Human artists/writers are influenced by each other all the time. I really don't see how it is fundamentally different. Most of Harry Potter is derivative of previous fantasy work itself. Nothing is made in a vacuum.

https://tvtropes.org/pmwiki/pmwiki.php/Main/WizardingSchool

Human artists/writers are influenced by each other all the time.

The flaw in this argument is the word "artist". If you remove all the pictures from the data source, the AI isnt capable of generating anything. Because it's not an artist.

Can a human "generate anything" beyond what essentially equates to random noise if they have never had any sensory input? Comparing a "trained" human brain with a "newborn" model seems strange if we actually want to delineate between what is and isn't art.

Ignoring the straw man argument here yes actually there are plenty of examples of individuals with no outside influence of art styles or references creating artworks. It's called outside art. https://en.wikipedia.org/wiki/Outsider_art

First, it would be great for you to explain how it is a strawman argument. In my eyes the strawman is comparing one system which has had years of training data and time with a system that has a rough structure, but misses everything beyond it. You'd have to at least give them comparable amounts of training data. Even a newly born baby has already had some sensory inputs in the womb, and starts to have a LOT afterwards.

Second, your Outsider Art is something completely different, funnily enough a strawman. Surely you're not claiming that the creators of outside art have literally never had any sensory input in their lives? Or do you really think that one painter in two timelines, in one blind and in the other not, would paint exactly the same stuff?

So if you were born blind, you can draw pictures?

I can assure you I absolutely could. Myself to start with - and possibly better than you could :)

Only a bodyless, artificial brain cant draw anything, blind.

Citation needed :)

But Rowling knew enough to pull from prior public domain works, not other recent authors. Wizard schools are public domain. An AI author would have to know which they are allowed to use, which they can use under fair use, and which they must ask to use. Humans can do that. I am doing that right now as I use the "Harry Potter" trademark here while posting to HN without the owner's permission. AI systems scraping the internet cannot understand that needed nuance.

Not everything JK pulled from is out of copyright. But generally Authors are allowed to reuse all sorts of things. Compare say Battle Royale to the Hunger Games. Writing a story about Kids fighting to the death doesn't give you a monopoly on that storyline.

See this link, lots of recent works in there.

https://en.wikipedia.org/wiki/Harry_Potter_influences_and_an...

Generally speaking though, any work created by such models does not copy any original work closely at all. Of course there could be slip-ups, but the same could be said of human generated works, which violate fair use on a regular basis.

And copyright law deals with the difference between inspiration and copying. To vastly oversimply it, it depends how close the original is to alleged copy.

No reason you can't apply that framework to AI.

Where AI might get into more trouble is that you might be able show literal copying in a way that it's impossible to do in a person mind. Like saving chunks of a work into its model.

Yes and when the derived things are very similar to the original and people feel wronged, they sue and judgements come from the court.

Now if you are the company selling this product, how many people are feeling wronged and will sue - that's the class action part?

If you use the product to generate an image that is very similar to someone's art and they feel wronged and sue, would you still use it commercially?

Humans aren't just reading, we're constantly updating our brains neural nets. Both the AI system and brains may be capable of writing a copyright infringing rip off of Harry Potter, but the ability to do so isn't infringement only actually doing so.

That wasn't the question. The question was if the learning process itself is violating any existing copyright laws.

That was part of the question, but it was immediately followed with a suggestion that things need to be virtually identical to be copyright violating.

Reply on news.ycombinator.com