Things are about to get worse for generative AI • Hacker News Board

Re: Things are about to get worse for generative AI

#391

Post by dangus » Sat, Dec 30, 2023, 3:50 PM UTC

madamelic wrote: post #329

Intox wrote: post #12

Or... things are about to get worse for copyright holders. I don't see any developped country pressing the brake on AGI in the near future to protect a few copyright holders from getting "stolen" in hypothetic scenarios.

Copyright should be the problem of the person using the works and not the problem of the AI generating it. Unless Nintendo plans on busting down the doors of every person who tries to draw Mario or preventing little Timmy from making a parody of Coca-Cola, making it where AI cannot generated copyrighted works is insane imo. Those brands should be proud to be such a big part of the cultural fabric that it is difficult…

The AI generating it (Hosted on OpenAI-controlled servers in the case of ChatGPT and DALL-E) is the entity redistributing the work. The end user who asked for the infringing content isn't the entity that is infringing on the copyrights and trademarks.

I'm perfectly free to ask people on the street for t-shirt with Mario on it, but as soon as someone who isn't Nintendo or licensed by Nintendo sells me that t-shirt they're the ones infringing on the copyright and trademark. As the consumer I did nothing illegal, and a court would say that I was deceived by the infringing party.

Distribution (seeding, uploading) and facilitating copyright infringement is what gets you in trouble. When you ask DALL-E (a paid, commercial product) for a picture of Italian plumbers and it gives you an obvious picture of Mario 100% recognizable to the layperson as Mario and not a distinctly different image of a similar character, that's blatant trademark and/or copyright infringement on the part of OpenAI.

> If the AI is completely unable to generate non-infringing works even if you are _trying_ to get away from it (which the author very much doesn't seem they are, they are purposefully making and show prompts that infringe), that's the problem of the AI creator then.

I see some parallels to the Napster lawsuit. The fact that the users were the bad people asking for infringing content didn't give Napster the right to facilitate infringement. Napster was ordered to monitor its network and make sure that they were blocking non-legitimate uses. They couldn't logistically comply and went bankrupt.

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

Which begs the question: Does OpenAI even have the technological ability to block trademark and copyright infringing content generation? Even if they do, how useful will ChatGPT be if all phrases and imagery that closely resemble copyrighted works are blocked from output?

Whats even worse for OpenAI compared to Napster is that it wasn’t individual users uploading copyrighted content, it was OpenAI’s ingesting the data. Nobody twisted their arm to include copyrighted works in their models.

Re: Things are about to get worse for generative AI

#392

Post by dartos » Sat, Dec 30, 2023, 3:53 PM UTC

tjr wrote: post #383

tsumnia wrote: post #354

Earlier quoted context omitted.

When I cover generative AI in my Ethics in AI lecture, one of few soapbox opinions I give is that GenAI is doing essentially what people do - copy others. Picasso has a quote about "Good Artists copy, Great Artists steal", which doesn't mean try to pass Lario and Muigi off as your own, but rather that great artists are able to take aspects from other works (also called 'inspiration') without being caught. My personal…

Timmy did not need to ingest the whole sum of human knowledge as a training set before he could draw a crayon copy of a cartoon character.

I mean… neither did any AI.

Re: Things are about to get worse for generative AI

#393

Post by kenmacd » Sat, Dec 30, 2023, 3:54 PM UTC

layer8 wrote: post #322

WhiteNoiz3 wrote: post #281

As I understood it, the legal precedent for generative AI is the same one that allows google to scrape websites in order to index them for search for the common good. Google also can display cached versions of websites which is the original content of those sites. No one is going to say that google is copyright infringement just because it is showing content from other websites verbatim. So I think this is a weak arg…

The ability to provide a reference to the source is the crucial difference here. I agree that it should be possible to implement that for generative AI, although the training may become significantly more expensive in order to maintain that information, and the AI companies have little interest in doing so. They’ll probably rather try to heuristically assess possible copyright issues after the fact in a post-processi…

> They’ll

What if I asked you to list all our source material that led you to use that particular contraction. Heuristics will not do, you must list each.

Can you do it? Do you believe AI should.

> I agree that it should be possible to implement

Those exact words appear in another forum post from 2006:

https://discourse.igniterealtime.org/t/cm-3beta-compression-...

Should you have quoted that as a source for your reply? What if we knew you'd read that post back in 2006, affecting your neurons, then should you?

It might not be too hard to imagine a simple case of a specific topic where you might have some more prominent sources, but even in those cases I believe if you think it through you'll find there was a ton of other sources that led to the weights that allowed you to 'know' the topic.

Re: Things are about to get worse for generative AI

#394

Post by sensanaty » Sat, Dec 30, 2023, 3:54 PM UTC

stale2002 wrote: post #379

schneems wrote: post #370

Earlier quoted context omitted.

Rounding up a transaction and taking the leftovers wouldn’t be a crime worthy of the FBI for one transaction but it would be for a million or a billion. Scale matters and impact matters. If you’re making an ethical argument “it’s okay because it’s already happening to a lesser degree somewhere else” isn’t the flex you think it is. If you’re talking ethics, talk about impact. Who does it help the most who does it hurt…

> “it’s okay because it’s already happening to a lesser degree somewhere else” isn’t the flex you think it is. It actually is. It shows that we as a society are completely OK with this, and nobody is complaining about a very standard and common thing that all artists do. It shows that the outrage is fake, and people don't actually care about the issue.

Do you really see no difference between someone drawing a piece of fan art and trillion dollar corporations stealing other people's works and reselling it for their own profit with no regards to anyone or anything else?

And yes, obviously society cares about many things depending on the scales in question. It's okay if a dude goes onto a lake on his small rowboat and catches a few fish for dinner, it's a completely different story if you're talking about a massive barge indiscriminately catching literally thousands of fish with huge nets. The latter has to adhere to much stricter rules than the prior, and I think you'd be hard pressed to find anyone who thinks these 2 situations should be treated equally (unless you're a commercial fisherman with a barge, I suppose, the quote "It is difficult to get a man to understand something when his salary depends on his not understanding it." comes to mind here)

Re: Things are about to get worse for generative AI

#395

Post by gaganyaan » Sat, Dec 30, 2023, 3:56 PM UTC

tjr wrote: post #383

tsumnia wrote: post #354

Earlier quoted context omitted.

When I cover generative AI in my Ethics in AI lecture, one of few soapbox opinions I give is that GenAI is doing essentially what people do - copy others. Picasso has a quote about "Good Artists copy, Great Artists steal", which doesn't mean try to pass Lario and Muigi off as your own, but rather that great artists are able to take aspects from other works (also called 'inspiration') without being caught. My personal…

Timmy did not need to ingest the whole sum of human knowledge as a training set before he could draw a crayon copy of a cartoon character.

Timmy did however spend years and years training on his own data ingested from incredibly high quality sensors. Not sure what your point is

Re: Things are about to get worse for generative AI

#396

Post by ben_w » Sat, Dec 30, 2023, 3:57 PM UTC

edgyquant wrote: post #380

ben_w wrote: post #229

Earlier quoted context omitted.

The reverse can also be true, with John Keats' agreeing with Charles Lamb that Newton "had destroyed all the poetry of the rainbow, by reducing it to the prismatic colours": https://en.wikipedia.org/wiki/Lamia_(poem) If we should ever fully understand how our own minds work, will we hold machines in higher esteem, or ourselves in lower?

Any biology or physics that suggests humans are just a pattern recognizer will be discarded as us being a conscious being is the only thing every human knows to be 100% true.

So, all of biology and physics then. If souls exist, they have no mass, and have a weird way of being repeatably disrupted in consistent ways by damage to certain parts of the brain or specific chemicals.

Just because consciousness is a mystery today, doesn't mean we get to stop and say it will be so forever more.

Heck, the problem still fundamentally exists regardless of if you're atheist, monotheist, polytheist, or pantheist.

--

“We’re not listening to you! You’re not even really alive!” said a priest.

Dorfl nodded. “This Is Fundamentally True,” he said.

“See? He admits it!”

“I Suggest You Take Me And Smash Me And Grind The Bits Into Fragments And Pound The Fragments Into Powder And Mill Them Again To The Finest Dust There Can Be, And I Believe You Will Not Find A Single Atom Of Life–”

“True! Let’s do it!”

“However, In Order To Test This Fully, One Of You Must Volunteer To Undergo The Same Process.”

There was silence.

“That’s not fair,” said a priest, after a while. “All anyone has to do is bake up your dust again and you’ll be alive…”

- Feet of Clay, Terry Pratchett

Re: Things are about to get worse for generative AI

#397

Post by rvz » Sat, Dec 30, 2023, 3:57 PM UTC

gumballindie wrote: post #377

ultrablack wrote: post #376

We are all trained on copyrighted input. That is not a problem. What is a problem is if you reproduce it and try to claim copyright for that. If someone wants to create their own image of Mario in an AI, so what?

We are not machines. The argument that procedural text and image generators are similar to us is ridiculous. The issue is not whether people can generate images. The issue is ai companies stealing content and reselling it. That needs to stop.

> The argument that procedural text and image generators are similar to us is ridiculous.

Agreed. The amount of endless whataboutisms AI proponents have to continuously invent around comparing humans and AI machines as having 'similar' characteristics to justify mass copyright violation is just absolutely laughable.

> The issue is ai companies stealing content and reselling it.

The key point here is the 'reselling' part, without credit, attribution or permission to do so and then claiming the creation as one's own. The fact that these AI companies won't disclose their training data, tells us that they know they are in deep trouble. The so-called 'fair use' excuses isn't going to work this time.

Given that Apple paid news orgs to train on their licensed data, the lawsuit with the NYT should not be a surprise for OpenAI and Microsoft (as they knew that they needed to pay for a license to access and train on the data) and will eventually end with a licensing deal with the NYT.

Re: Things are about to get worse for generative AI

#398

Post by sensanaty » Sat, Dec 30, 2023, 3:58 PM UTC

dartos wrote: post #392

tjr wrote: post #383

Earlier quoted context omitted.

Timmy did not need to ingest the whole sum of human knowledge as a training set before he could draw a crayon copy of a cartoon character.

I mean… neither did any AI.

Wasn't ChatGPT trained on the entirety of Wikipedia? And probably millions of pieces of scientific literature, and arts, and movies and games and and and...

Perhaps the hyperbole of the entire corpus of human knowledge isn't quite technically right, but it's close enough.

Re: Things are about to get worse for generative AI

#399

Post by firtoz » Sat, Dec 30, 2023, 3:58 PM UTC

tjr wrote: post #383

tsumnia wrote: post #354

Earlier quoted context omitted.

When I cover generative AI in my Ethics in AI lecture, one of few soapbox opinions I give is that GenAI is doing essentially what people do - copy others. Picasso has a quote about "Good Artists copy, Great Artists steal", which doesn't mean try to pass Lario and Muigi off as your own, but rather that great artists are able to take aspects from other works (also called 'inspiration') without being caught. My personal…

Timmy did not need to ingest the whole sum of human knowledge as a training set before he could draw a crayon copy of a cartoon character.

He does ingest a lot of frames of input consisting of mostly that.

Re: Things are about to get worse for generative AI

#400

Post by TheRoque » Sat, Dec 30, 2023, 3:59 PM UTC

keiferski wrote: post #39

These don't seem all that difficult to fix to me. Most of the examples are not really generic, but are shorthand descriptions of well-known entities. "Video game plumber" is practically synonymous with "Mario" and anyone that has the slightest familiarity with the character knows this. Likewise, how difficult is it to just use descriptive tools to describe Mario-like images [1] and then remove these results from anyo…

How do you know you are inputing "well known entities" if you don't know it beforehand ? If I type "columbian coffee logo" and end up with logos of brands that existed before hand, should I just reverse engineer the whole internet to find out if these logos existed already or not ? The AI should show its inspiration. A human who takes inspiration of something else for its creation knows precisely what it used, and if he crossed the line of plagiarism or not, but the way AI work are too opaque for that. I think the thing it needs to do is reveal its sources, nothing more, but it also means for the AI companies to reveal their dataset, and maybe information they shouldn't have, nor disclose.