AI is stealing writers’ words and jobs…

[trappedslider]

Am I doing this right?

Actual headline should read "evil stealing bad scientist use unpaid artwork to make model faster"

New AI image generator is 8 times faster than OpenAI's best tool — and can run on cheap computers

Scientists used "knowledge distillation" to condense Stable Diffusion XL into a much leaner, more efficient AI image generation model that can run on low-cost hardware.

www.livescience.com

 

[Scribe]

Am I doing this right?

Actual headline should read "evil stealing bad scientist use unpaid artwork to make model faster"

New AI image generator is 8 times faster than OpenAI's best tool — and can run on cheap computers

Scientists used "knowledge distillation" to condense Stable Diffusion XL into a much leaner, more efficient AI image generation model that can run on low-cost hardware.

www.livescience.com

I mean does it address actual issues or is it just more 'rah rah "progress" good yeah!'?

 

[trappedslider]

I mean does it address actual issues or is it just more 'rah rah "progress" good yeah!'?

depends on what you feel is the issue. For some the issue is not being able to run models locally but for the audience here the issue is data scraping (gross simplification but it is what it is).

__

China court says AI broke copyright law in apparent world first | Semafor

A Chinese court ruled that an AI bot improperly used protected material in generating images in what appears to be the first ruling of its kind.

www.semafor.com

__

OpenAI alleges New York Times 'hacked' ChatGPT for lawsuit evidence

OpenAI asked a judge to dismiss parts of The New York Times' lawsuit against it, alleging that the media company "paid someone to hack OpenAI's products."

www.cnbc.com

 

[Ryujin]

China has some seriously draconian copyright and trademark laws, with severe penalties. Pity that they aren't frequently used.

I used to watch the dubbed version of the original Ultraman on WUTV, out of Grand Island, NY.

 

[trappedslider]

China has some seriously draconian copyright and trademark laws, with severe penalties. Pity that they aren't frequently used.

In the Beijing case, the court found that generative AI is “merely a tool that assists the plaintiff in his creation” of art, while the Guangzhou ruling determined that AI “participated in the creation of content involved in the case, rather than being purely instrumental,”

so basically just like US law, they have no idea how to handle it. Which is basically how tech works now, first tech happens then come the laws.

 

[nevin]

so the big issue now is the AI bots (though smart scripting bot's is a better term. they aren't really AI) were all trained using data scraped from the internet using Fair Use Doctrine. Now the issue is the Bots aren't smart enough to not violate copyright law or other intellectual property laws and some companies are taking advantage, tech companies are playing a slow defensive action to prevent fair use law from being taken from the AI firms who need the large data sets to program them. Techs been doing stuff like this to the blue collar worker's for decades. I think they were caught by surprise when famous artists, writers and hollywood started going after them.

 

[Jfdlsjfd]

so the big issue now is the AI bots (though smart scripting bot's is a better term. they aren't really AI) were all trained using data scraped from the internet using Fair Use Doctrine.

No, the smart scripted bots wouldn't need training nor data scraped from anywhere. That's the difference with AI.

Now the issue is the Bots aren't smart enough to not violate copyright law or other intellectual property laws and some companies are taking advantage, tech companies are playing a slow defensive action to prevent fair use law from being taken from the AI firms who need the large data sets to program them.

They don't really need the large data sets. If large datasets are freely available, then using them is cheaper than the best alternative, smaller but relevant datasets. If they can no longer use free large datasets by scraping, which they will (the TDM exception in the EU only requires respecting opt-out -- it will still be cheaper to train in a EU datacenter than buying data, and Singapore is implementing even laxer restrictions), the next best alternative isn't paying for data. For generative AI, It's automatically captioning of a smaller dataset and potentially have a low-wage validation of the captioning if needed.

Techs been doing stuff like this to the blue collar worker's for decades. I think they were caught by surprise when famous artists, writers and hollywood started going after them.

Thinking that multi-billions companies are caught by surprise by the concept of a lawsuit happening in the US is certainly giving them very little credit. I can't imagine Microsoft legal team not anticipating a lawsuit happening.

 

[nevin]

From Science Daily.

"The people who made LLMs call it "hallucinating" when they make things up; although Chemero says, "it would be better to call it 'bullsh*tting,'" because LLMs just make sentences by repeatedly adding the most statistically likely next word -- and they don't know or care whether what they say is true."

AI is not intelligent. It pulls the most statistically relevant information to create something that statistically matches the question. It's smart scripting, it's not intelligence. AI was the absolute worst name to use to explain what it does. But it was a great marketing tool to get people excited. it's more like the millionth step in a billion to getting to actual Artificiall Intelligence.

 

[Jfdlsjfd]

Still, going by a script is different from what it does. Using statistically relevent information is different from running a script with predefined possibilities at each interaction. I thought you were speaking about a subset of uses (like in the earliest chatbot) but you're just proposing to call AI, instead of the commonly used name and well defined (nobody think there is a little guy thinking in the computer, sure it's a metaphor instead of a very accurate depiction, but we call hand-held 2-dimensional pointing device mice, so it's kind of common practice), by an even more confusing name.

 

[trappedslider]

OMG another slightly less evil corporation is wanting money by selling out it's user base (what's left of it) to another more evil corporation

Tumblr

Tumblr is a place to express yourself, discover yourself, and bond over the stuff you love. It's where your interests connect you with your people.

www.tumblr.com

to save you form having to click on the line here's the post :

Tumblr
jv

Follow


Anonymous asked:
What is this about the tumblr staff wanting to sell art data to midjourney?


An ex-colleague of mine mentioned yesterday that there may be contacts between Automattic and midjourney in that direction, but nothing is public yet and I don't have any more info. They probably won't have anything specific to share either, since they left the company weeks ago too. That being said:

• I have no reason to doubt my ex-coworker word, they are a trustworthy person.
• Tumblr's CEO has been absurdly enthusiastic (comically, even) about AI, and is a big fan of LLMs and 'AI' companies.
• A deal with midjourney could solve tumblr financial issues (not the same company, but openAi is paying up to 5 million/year to news companies to use their content as training data... tumblr generates several orders of magnitude more content than any newspaper or any media company and it only would need a 20 to 30 million per year deal to be profitable)

So I don't have any extra info yet, but I'm keeping my ears open.

Actual news on it Tumblr and WordPress Are Selling Your Data to AI Companies