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

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.

VanguardLH wrote:
> 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.

The relatively simple parallel case of informing callers that their talk
may be recorded for staff training etc. has its complications. In the UK
it's not a legal requirement, just considered polite. But in the USA
different states have different stances on the question. Take that
proposition into the WWW realm and you'll tumble into a whole bag of
worms. California alone has seen many recent such wrangles.

Ed

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 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.

--
....w¡ñ§±¤ñ

...winston 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.
>
Don't forget how Google pixelled out faces on Google Earth. They did
that for a legal reason, not a humanitarian one.
Ed

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.

VanguardLH <V@nguard.lh> wrote:
> 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.

False. You're just confirming the misconception. As I said,
'copyright' is implicit and default. The one contending it has to
provide proof to the contrary, show earlier work, etc.. It does *not*
have to have special quality. Even a shopping list is 'copyright'ed.
Note that - as I said - am talking about private people/individuals,
*not* about businesses, corporations, organizations, etc..

> > 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.

In international law, there is no such thing as "the public domain".
It may exist in some *local* laws, but then is only applicable to *that*
jurisdiction. And - again - 'copyright' is the default, one does not
have to to add a copyright statement.

Actually, in international law, one can not even put something in "the
public domain". One can make something free to use, free to copy, free
to change (but referring to the original), etc., but one still has some
authors rights (for example the requirement to attribute).

Again, local law might be more limited, but only applies to the local
jurisdisction.

It's probably because of the 'copyright' misnomer and the fact that US
law is more limited, that people, especially Americans, think that their
situation implies worldwide. It doesn't.

> > 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.

They don't have to be defended. If someone doesn't want to accept the
copyright, *they* need to challenge it

> > 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.

Sigh! Of course "can not" is shorthand for "is not allowed to". Of
course one *can* violate authors rights, but that doesn't make it legal.

Case in point: Quite some time ago, someone was getting postings from
several nl.* groups and put those postings on a website of his, but he
didn't attribute those postings and changed them, IIRC mainly by not
showing them in full and without snip-marks. (Yes, he was quite a loon.)

We urged him to stop that or undo the violations, pointing him to the
relevant authors rights. It took a threat with a cease-and-desist order
to let him back down.

....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"! :-)

VanguardLH <V@nguard.LH> wrote:
> 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.

Any scientific study requires ethical approval in order to be able to use
human subjects and/or human data. That's normal practice.

> 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?

Is it personally identifiable data? Nope. So no.

> ""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.

No need to prove anything. Copyright is implicit for any creative work. So
any image or text not explicitly with an open licence will likely have been
infringed.

>> 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.

Personal data is very clearly defined in most countries.

Frank Slootweg wrote:
> ...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"! :-)
>
It can be a grey area.
I noted the U.S. as an example. Even 'International' copyright law has
its constraints
- i.e. country specific where protection does/does not apply to
multiple locations. Also cross-country treaties may/may not be in place.
Thus, applicability may be unique, not globally across locations.

--
....w¡ñ§±¤ñ

"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. Just as with a book. You can
draw in it, tear it apart, give it away to someone else,
sell it, quote from it... the only thing you can't do is
distribute copies.

Ironically, software companies and streaming services want
you to think that you don't have those rights with copyrighted
products. They want you to believe that their software is
"licensed" and that movies are broadcasts rather than
downloads. Before MS started with their rental plan they
were claiming that Windows is co-licensed to you and the
plastic entity known as your motherboard. But the bottom line,
despite MS's illegal restrictions, is that if you have a copy of
Windows then you have a right to use it or sell it or give it
away; and if you visit a website then you're legally obtaining
copies of those files. So if you wanted to claim illegal activity
then you'd have to claim that no one has a right to create a
product that involves reading data in those files.

And don't
forget that people have sued Google for rifling through their
emails to gmail users, in order to target ads, without the
sender's permission, and those lawsuits have been thrown out.
(In at least one case Google offered the defense that the sender
could reasonably be expected to know about the spying
because everyone knows that Google is a sleazeball operation.
Thus, the sender gave Google their private letters willingly. :)

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.

Newyana2 <Newyana2@invalid.nospam> wrote:
> "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.

As I mentioned, I was just responding to VanguardLH's (common)
misconception about 'copyright'.

It does, at least not directly apply, to the original topic of the
thread ("OpenAI, Microsoft face class-action suit over internet data use
for AI models").

[...]

> 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.

I'm not aware of such (Gmail scraping) an European/EU case against
Google. But I'm sure there are other cases against Google and the
others.

These days, I'm losing track of who is the bigtech-of-the-week that
gets in trouble with/in the EU. They all get slapped one time or
another, Google, Microsoft, Apple, Facebook, etc., etc..

> MS processing website content
> is arguably far less intrusive because the content has been
> specifically made public.

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.

On 7/18/2023 8:30 AM, Chris wrote:
> 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.
>
>
>

Unless the developers can claim fair use/fair dealing because they were
inspiring the AI/LLM to create works like those, but not verbatim
copying. If anything this will lead to developers training off freely
licensed works per say Kevin McLeod.