Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial intelligence tech like chatbot ChatGPT.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
The 157-page lawsuit written by Ryan Clarkson, ...
"Once trained on stolen data, defendants saw the immediate profit
potential and rushed the products to market without implementing proper safeguards or controls to ensure that they would not produce or support harmful or malicious content and conduct that could further violate the
law, infringe rights and endanger lives," Clarkson continued. "Without
these safeguards, the products have already demonstrated their ability
to harm humans, in real ways."
https://www.foxnews.com/tech/openai-microsoft-face-class-action-suit-internet-data-use-ai-models
Johnny <johnny@invalid.net> wrote:
Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of
California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial
intelligence tech like chatbot ChatGPT.
The computer community knows that Windows Home editions are beta
versions using Microsoft's consumers to test the product.
How do you gather information on human behavior without actually
monitoring human behavior? Hmm, guess the lawyers should also include
all the psychiatrists that based their analysis on all the recorded
history of human nature.
Just because someone opened a lawsuit does not mandate there is
foundation for their lawsuit. I suspect this one will drag out for 1 or
2 years, and then the judge will discard it as frivilous.
The lawyers don't have anyone paying them, so, yeah, start a lawsuit to
get some money that way. Ambulance chasers.
When checking how many pedestrians use which side of a two-way road correlates to the direction of automobile traffic on the same side of
the road, do you really have to stop all those pedestrians asking their permission to watch them walking down a sidewalk?
""unprecedented theft of private and copyrighted information belonging
to real people, ..."
Well, that's a lot to prove. Prove just what private information was
stolen, and that those watched didn't overty give permissions (or inadvertently grant permission by not bothering to read the TOS). Prove copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted material.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
"Personal data" will need to be excrutiatingly defined for this lawsuit
to have a snowball's chance in hell of not getting dismissed.
The 157-page lawsuit written by Ryan Clarkson, ...
An old tactic of lawyers: toss tons of documentation at the judge as
though the weightier the documentation the more it just must prove the
case.
"Once trained on stolen data, defendants saw the immediate profit
potential and rushed the products to market without implementing proper
safeguards or controls to ensure that they would not produce or support
harmful or malicious content and conduct that could further violate the
law, infringe rights and endanger lives," Clarkson continued. "Without
these safeguards, the products have already demonstrated their ability
to harm humans, in real ways."
Oh, the argument of what could be instead of what is.
Lawyers hate to be idle.
https://www.foxnews.com/tech/openai-microsoft-face-class-action-suit-internet-data-use-ai-models
"AI will probably most likely lead to the end of the world"
They need to stop watching televison and movies.
Prove
copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted material.
VanguardLH <V@nguard.lh> wrote:
[...]
[Just addressing this point:]
Prove
copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted
material.
Common misconception: Private people's material is copyrighted by
default. Nothing has to be done. International law (the Berne
Convention) says so.
So, for example, this posting is copyrighted.
Sadly, 'copyright' is a misnomer, because it sort of implies it's only about the right (not) to copy, which is false. That's why the Berne Convention and most local laws (including our Dutch one) talk about
authors (or creators) rights. (Wikipedia: "In some jurisdictions these
type of rights are being referred to as copyright.")
For example this posting can be and obviously is copied, because I
gave implicit permission to do so, but I still have the authors rights,
which for example means that my text can not be used without
attribution, can not be changed, etc., etc..
[...]
VanguardLH <V@nguard.lh> wrote:
[...]
[Just addressing this point:]
Prove
copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted
material.
Common misconception: Private people's material is copyrighted by
default. Nothing has to be done. International law (the Berne
Convention) says so.
So, for example, this posting is copyrighted.
Sadly, 'copyright' is a misnomer, because it sort of implies it's only about the right (not) to copy, which is false. That's why the Berne Convention and most local laws (including our Dutch one) talk about
authors (or creators) rights. (Wikipedia: "In some jurisdictions these
type of rights are being referred to as copyright.")
For example this posting can be and obviously is copied, because I
gave implicit permission to do so, but I still have the authors rights,
which for example means that my text can not be used without
attribution, can not be changed, etc., etc..
Frank Slootweg <this@ddress.is.invalid> wrote:
VanguardLH <V@nguard.lh> wrote:
[...]
[Just addressing this point:]
Prove
copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted >> material.
Common misconception: Private people's material is copyrighted by default. Nothing has to be done. International law (the Berne
Convention) says so.
Another misconception. Yes, your private material can be considered copyrighted, but you'll have to prove it in court. You have to show the material was generated before someone else's, and that it has qualities
that make it copyrightable. Assumed copyright usually gets discarded
without adequate proof of ownership.
So, for example, this posting is copyrighted.
Not since it was published in the public domain, and you posted no
copyright statement, and because then communications venue obviates copyrighting.
Sadly, 'copyright' is a misnomer, because it sort of implies it's only about the right (not) to copy, which is false. That's why the Berne Convention and most local laws (including our Dutch one) talk about
authors (or creators) rights. (Wikipedia: "In some jurisdictions these
type of rights are being referred to as copyright.")
Quite true. But when the lawyers get involved, anything regarding
copyright has to be registered. Assumed copyrights are not defendable.
For example this posting can be and obviously is copied, because I
gave implicit permission to do so, but I still have the authors rights, which for example means that my text can not be used without
attribution, can not be changed, etc., etc..
Good luck with that.
Frank Slootweg wrote:
VanguardLH <V@nguard.lh> wrote:
[...]
[Just addressing this point:]
Prove
copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted >> material.
Common misconception: Private people's material is copyrighted by default. Nothing has to be done. International law (the Berne
Convention) says so.
So, for example, this posting is copyrighted.
Sadly, 'copyright' is a misnomer, because it sort of implies it's only about the right (not) to copy, which is false. That's why the Berne Convention and most local laws (including our Dutch one) talk about
authors (or creators) rights. (Wikipedia: "In some jurisdictions these
type of rights are being referred to as copyright.")
For example this posting can be and obviously is copied, because I
gave implicit permission to do so, but I still have the authors rights, which for example means that my text can not be used without
attribution, can not be changed, etc., etc..
[...]
Just an fyi.
In the U.S. facts or a collection of facts are not protected by
copyright law.
i.e. a post to a public accessible server that provides a
factual(implicit, intended or supposedly) opinion may not always
guarantee copyright protection.
Johnny <johnny@invalid.net> wrote:
Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of
California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial
intelligence tech like chatbot ChatGPT.
The computer community knows that Windows Home editions are beta
versions using Microsoft's consumers to test the product.
How do you gather information on human behavior without actually
monitoring human behavior? Hmm, guess the lawyers should also include
all the psychiatrists that based their analysis on all the recorded
history of human nature.
Just because someone opened a lawsuit does not mandate there is
foundation for their lawsuit. I suspect this one will drag out for 1 or
2 years, and then the judge will discard it as frivilous.
The lawyers don't have anyone paying them, so, yeah, start a lawsuit to
get some money that way. Ambulance chasers.
When checking how many pedestrians use which side of a two-way road correlates to the direction of automobile traffic on the same side of
the road, do you really have to stop all those pedestrians asking their permission to watch them walking down a sidewalk?
""unprecedented theft of private and copyrighted information belonging
to real people, ..."
Well, that's a lot to prove. Prove just what private information was
stolen, and that those watched didn't overty give permissions (or inadvertently grant permission by not bothering to read the TOS). Prove copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted material.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
"Personal data" will need to be excrutiatingly defined for this lawsuit
to have a snowball's chance in hell of not getting dismissed.
...winston <winstonmvp@gmail.com> wrote:
Frank Slootweg wrote:
VanguardLH <V@nguard.lh> wrote:
[...]
[Just addressing this point:]
Prove >>>> copyright infringement which can be very tricky if all those
participants didn't bother to actually copyright the claimed copyrighted >>>> material.
Common misconception: Private people's material is copyrighted by
default. Nothing has to be done. International law (the Berne
Convention) says so.
So, for example, this posting is copyrighted.
Sadly, 'copyright' is a misnomer, because it sort of implies it's only >>> about the right (not) to copy, which is false. That's why the Berne
Convention and most local laws (including our Dutch one) talk about
authors (or creators) rights. (Wikipedia: "In some jurisdictions these
type of rights are being referred to as copyright.")
For example this posting can be and obviously is copied, because I
gave implicit permission to do so, but I still have the authors rights,
which for example means that my text can not be used without
attribution, can not be changed, etc., etc..
[...]
Just an fyi.
In the U.S. facts or a collection of facts are not protected by
copyright law.
i.e. a post to a public accessible server that provides a
factual(implicit, intended or supposedly) opinion may not always
guarantee copyright protection.
I take your word [1] for it. And, as you rightly say, this applies to
"In the U.S.", i.e. within the US jurisdiction(s).
What I'm referring to, is international 'copyright' law, i.e. between countries. International 'copyright' law says that my posting is 'copyrighted' by default and hence you can not (read: are not allowed
to) violate the mentioned authors rights (attribution, change).
So, since you *did* attribute and quoted in full, you're probably
safe! :-)
[1] Before I saw my Freudian slip, the word was "work"! :-)
"Frank Slootweg" <this@ddress.is.invalid> wrote
| What I'm referring to, is international 'copyright' law, i.e. between
| countries. International 'copyright' law says that my posting is
| 'copyrighted' by default and hence you can not (read: are not allowed
| to) violate the mentioned authors rights (attribution, change).
|
I don't see how that applies. Microsoft don't seem to
be quoting sections of online text. They're scraping
websites for data to train their bot. If you put up a
website, by the design of http, you're inviting the world
to copy your files. The only illegal thing would be to copy
and distribute those files.
That seems to be a case very similar to ChatGPT. Google is
processing the emails with software but not publicing or re-using
the content. It's crazy to me that that's legal, but that's how
the courts have decided it. I'm not aware of any similar lawsuits
succeeding in Europe. Is there such a case? I'm only aware of
fines for companies storing Euro data in the US. Presumably
Google rifles through "googlemail" email in the EU, just as it
rifles through gmail email in the US.
MS processing website content
is arguably far less intrusive because the content has been
specifically made public.
Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial intelligence tech like chatbot ChatGPT.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
The 157-page lawsuit written by Ryan Clarkson, the managing partner of
the firm, also asserts that without the "unprecedented theft of private
and copyrighted information belonging to real people," the products
developed by the companies "would not be the multi-billion-dollar
business they are today."
"Once trained on stolen data, defendants saw the immediate profit
potential and rushed the products to market without implementing proper safeguards or controls to ensure that they would not produce or support harmful or malicious content and conduct that could further violate the
law, infringe rights and endanger lives," Clarkson continued. "Without
these safeguards, the products have already demonstrated their ability
to harm humans, in real ways."
https://www.foxnews.com/tech/openai-microsoft-face-class-action-suit-internet-data-use-ai-models
On 6/29/2023 4:17 PM, Johnny wrote:
Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of
California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial
intelligence tech like chatbot ChatGPT.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
The 157-page lawsuit written by Ryan Clarkson, the managing partner of
the firm, also asserts that without the "unprecedented theft of private
and copyrighted information belonging to real people," the products
developed by the companies "would not be the multi-billion-dollar
business they are today."
"Once trained on stolen data, defendants saw the immediate profit
potential and rushed the products to market without implementing proper
safeguards or controls to ensure that they would not produce or support
harmful or malicious content and conduct that could further violate the
law, infringe rights and endanger lives," Clarkson continued. "Without
these safeguards, the products have already demonstrated their ability
to harm humans, in real ways."
https://www.foxnews.com/tech/openai-microsoft-face-class-action-suit-internet-data-use-ai-models
machine learning is no different than humans learning off stuff. If you
show me a picture, there is potential copyright infringement because if
I create a picture I could unconsciously infringe.
LLMs (Fake AIs) don't
have a conscious
and therefore cant infringe on copyrights.
josh allen <josha12@googlemail.com> wrote:
On 6/29/2023 4:17 PM, Johnny wrote:
Published June 29, 2023
A class-action complaint filed Wednesday in the northern district of
California alleges tech leaders OpenAI and Microsoft Corp. used "stolen
and misappropriated" information from hundreds of millions of internet
users without their knowledge to train and develop its artificial
intelligence tech like chatbot ChatGPT.
The 16 plaintiffs, who are represented by the Clarkson Law Firm and
listed with initials, claimed the defendants "continue to unlawfully
collect and feed additional personal data from millions" worldwide to
that end and that they systematically scraped 300 billion words from
the internet without consent.
The 157-page lawsuit written by Ryan Clarkson, the managing partner of
the firm, also asserts that without the "unprecedented theft of private
and copyrighted information belonging to real people," the products
developed by the companies "would not be the multi-billion-dollar
business they are today."
"Once trained on stolen data, defendants saw the immediate profit
potential and rushed the products to market without implementing proper
safeguards or controls to ensure that they would not produce or support
harmful or malicious content and conduct that could further violate the
law, infringe rights and endanger lives," Clarkson continued. "Without
these safeguards, the products have already demonstrated their ability
to harm humans, in real ways."
https://www.foxnews.com/tech/openai-microsoft-face-class-action-suit-internet-data-use-ai-models
machine learning is no different than humans learning off stuff. If you
show me a picture, there is potential copyright infringement because if
I create a picture I could unconsciously infringe.
Correct. Which is why there's many lawsuits between musicians alleging
such.
LLMs (Fake AIs) don't
have a conscious
A conscious what?
and therefore cant infringe on copyrights.
Lol. No-one is claiming the LLM itself is infringing, but the developers of the LLM can be held accountable regarding what data they used to train the LLM.
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 297 |
Nodes: | 16 (2 / 14) |
Uptime: | 08:10:34 |
Calls: | 6,666 |
Files: | 12,213 |
Messages: | 5,336,192 |