California bill (AB 412) would effectively ban open-source generative AI

[trappedslider]

Least we forget that the web scrapping is how the datasets are being created, which despite your dislike is currently still legal under the right circumstances as far as image generators go.

Kneschke vs. LAION - Landmark Ruling on TDM exceptions for AI training data – Part 1 - Kluwer Copyright Blog

On September 27, 2024, the German Hamburg Regional Court (“Court”) issued the first ruling on reproductions of copyrighted content from the Internet made during the creation of an AI training data set – and on whether the copyright exceptions for text and data mining (“TDM”) provide statutory...

copyrightblog.kluweriplaw.com

Kneschke vs. LAION - Landmark Ruling on TDM exceptions for AI training data – Part 2 - Kluwer Copyright Blog

In the first part of this post on the Kneschke vs. LAION decision by the German Hamburg Regional Court (“Court”), we explored the Court’s key findings regarding the operational step in a generative AI model, and the decision on the exceptions for scientific research text and data mining (“TDM”)...

copyrightblog.kluweriplaw.com

and the growing number of websites that are selling the permissions to scrap for data, reddit, a number of forums,

A poster’s guide to who’s selling your data to train AI

Those Tumblr, Reddit, and WordPress posts you never thought would see the light of day? Yep, them too.

www.vox.com

so at this point, it honestly seems the only way NOT to get swept up and is pay a company to find every byte of data about you on the web, every post you and have it erased. (My inner greedy goblin now has me wondering how much the data from here is worth price wise) also proving @Scribe right that all developments made after the mid 90s (widespread mailing of AOL disc) in the area of the internet was a vast mistake.

 

[Scribe]

proving @Scribe right

 

[Jfdlsjfd]

Least we forget that the web scrapping is how the datasets are being created, which despite your dislike is currently still legal under the right circumstances as far as image generators go.

Yes, it is one of the way to resolve the problem with the datasets. Much like the State limits the scope of physical ownership for the common good (depending on where you are, you can't own air, the subsoil, other people, game...), some can choose to limit the scope of intellectual ownership for the common good (so you can't own the right to oppose AI training, or the right to oppose parody, and so on...).

and the growing number of websites that are selling the permissions to scrap for data, reddit, a number of forums,

A poster’s guide to who’s selling your data to train AI

Those Tumblr, Reddit, and WordPress posts you never thought would see the light of day? Yep, them too.

www.vox.com

Yes. The amount of images available for training (from various source, even if one were to not live in a place where scraping is allowed):

• Licensed through hosting websites,
• Public domain works,
• Database of images specifically made under permissive licenses,
• Images distribued from public learning institutions...

Is enough to train models without resorting to image datamining. One of the most probable outcome of the opposition to data scraping is that models trained on non-scraped content will progressively replace everything else in countries that don't allow scraping, so anyone will be able o get AI images from nearly free but based on the aforementioned content as a substitute for scraped contents. It will have no net effect on artists, except that instead of being replaced by AI trained on their work, they will be replaced by AI trained on other's work.

the mid 90s (widespread mailing of AOL disc) in the area of the internet was a vast mistake.

I remember that, they nearly replaced drink coasters at the time!

 

[Blue]

Yeah that could drive the price up so high it could become as overpriced as human writers and artists, completely completing the cost-cutting benefit of this technology and making bespoke writing and artwork once again the exclusive purview of the rich elite who can afford to throw their money away.

I can't afford to comission an artist! I have car payments and medical bills and a crappy low wage job! But I can easily afford a subscription to artbreeder and NovelAI. Why should I be denied that just because I'm not rich

Why should you be denied a private jet just because you weren't rich?

Same concept - people have put effort into creating something, why do y9ou think they shouldn't be compensated?

This comes off as very entitled, that just because you can't afford something, the people who make it shouldn't get paid for their work so you can have it for free.

 

[Blue]

Most of the stuff that's copyrighted now should be public domain but the big media companies keep bribing congress to extend the duration of copyrights. If it lasted a reasonable amount of time like patents do I'd be with you, but not under the farcical current laws where copyrights can easily last two entire lifetimes

Citation please.

There is so much art created during the digital era, it dwarfs what has been put online from previous art. There are DeviantArt users that have more images up than some Museums. And let me tell you, there's a lot more users than museums.

 

[trappedslider]

Citation please.

There is so much art created during the digital era, it dwarfs what has been put online from previous art. There are DeviantArt users that have more images up than some Museums. And let me tell you, there's a lot more users than museums.

This was the big part of the uproar over DMCA back in '98 Copyright Term Extension Act - Wikipedia

Copyright law of the United States - Wikipedia

en.wikipedia.org

I'll note that Deviant has taken the side of the users but they also note that : In actuality, an AI generator looks at millions of images to understand shapes, patterns, and colors – similar to how an artist might search the web for photos of a tiger before drawing their own tiger – and then creates a brand-new image. They have also put in their own generator so

 

[trappedslider]

I almost said something about scrubbing your digital footprint before you disappear into the back country

 

[Scribe]

I almost said something about scrubbing your digital footprint before you disappear into the back country

I won't care by then, I don't particularly care about that now.

I cannot imagine what it would take to scrub everything I've said over the last .... yikes it's 2025...

 

[trappedslider]

I won't care by then, I don't particularly care about that now.

I cannot imagine what it would take to scrub everything I've said over the last .... yikes it's 2025...

money/time

1. Digital Privacy Services: Companies like DeleteMe, PrivacyDuck, and Reputation Defender offer services to help individuals remove their personal information from data broker websites, search engines, and social media platforms.
2. Cybersecurity Firms: Some cybersecurity firms provide comprehensive digital footprint management, including monitoring and removal services. Look for firms that specialize in personal privacy.
3. Freelancers: You can hire freelancers with skills in digital privacy and cybersecurity through platforms like Upwork or Fiverr. Look for profiles that highlight experience in digital footprint removal.
4. Legal Services: In some cases, consulting with a lawyer who specializes in privacy law can be beneficial, especially if you need to address specific legal issues related to your digital footprint.
5. DIY Approaches: While hiring professionals can be effective, you can also take steps yourself, such as:
   • Deleting unused social media accounts.
   • Opting out of data broker sites.
   • Adjusting privacy settings on existing accounts.
   • Requesting removal of personal information from search engines.

 

[gnarlygninja]

This was the big part of the uproar over DMCA back in '98 Copyright Term Extension Act - Wikipedia

Copyright law of the United States - Wikipedia

en.wikipedia.org

I'll note that Deviant has taken the side of the users but they also note that : In actuality, an AI generator looks at millions of images to understand shapes, patterns, and colors – similar to how an artist might search the web for photos of a tiger before drawing their own tiger – and then creates a brand-new image. They have also put in their own generator so

I think they were specifically talking about the "most stuff under copyright should be in the public domain" part, not the part about companies continuously extending the duration (although the last time they tried, it failed)