to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without contravening
the 1st amendment because it was written by an AI engine.

Corollary: Would it be possible to ban a random string of numbers, as
they do not make up intelligible "speech" ?

And as a final kicker, could a state of affairs arise where AI output is considered "not speech" (and therefore subject to bans) whilst
simultaneously being "speech" (and thus worthy of copyright and the
financial implications that brings) ?

If no one has asked these questions before, I find myself idly wondering
what a country with so many lawyers such as the US is doing with it's
time. Surely there aren't that many pictures of statues to occupy your
greatest minds ? (I write as a UKian watching from afar).

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 29/03/23 19:49, Jethro_uk wrote:

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without contravening
the 1st amendment because it was written by an AI engine.

Corollary: Would it be possible to ban a random string of numbers, as
they do not make up intelligible "speech" ?

And as a final kicker, could a state of affairs arise where AI output is considered "not speech" (and therefore subject to bans) whilst
simultaneously being "speech" (and thus worthy of copyright and the
financial implications that brings) ?

If no one has asked these questions before, I find myself idly wondering
what a country with so many lawyers such as the US is doing with it's
time. Surely there aren't that many pictures of statues to occupy your greatest minds ? (I write as a UKian watching from afar).

If the 1ˢᵗ amendment is not extended to non-homo sapiens, let alone non-Americans,
why would it be extended to AI, at large? Its developer could claim it's autonomously
acting on behalf of him/her. But a true, unowned autonomous AI would not have 1ˢᵗ amendment rights.

IANAL.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Jethro_uk <jethro_uk@hotmailbin.com> wrote:

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without
contravening the 1st amendment because it was written by an AI
engine.

Well, a corporation is not a human being, but the Supreme Court has
ruled that a corporation is entitled to free speech rights. Not only
that but money is essentially considered speech. So that if a
corporation wants to spend millions of dollars to sway voters, while
ordinary citizens don't have that kind of money, the corporation is
only exercising its free speech rights.

A female friend once got exasperated about the rights of women in the
US these days, and said, "women aren't people." To which I properly
responded, "then women should incorporate."

Corollary: Would it be possible to ban a random string of numbers,
as they do not make up intelligible "speech" ?

Speech doesn't have to be meaningful, or even intelligible to be
protected.

And as a final kicker, could a state of affairs arise where AI
output is considered "not speech" (and therefore subject to bans)
whilst simultaneously being "speech" (and thus worthy of copyright
and the financial implications that brings) ?

Well, copyright's a whole other issue. I suppose your hypothetical
is possible - it's really up to the courts to sort that kind of thing
out. There's no good way to tell in advance what they mighgt do in
an unusual situation like that.

--
Stu
http://DownToEarthLawyer.com


--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

According to Jethro_uk <jethro_uk@hotmailbin.com>:

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without contravening
the 1st amendment because it was written by an AI engine.

The Supreme Court will shortly rule on Gonzalez v. Google where the argument
is whether Youtube's automated recommendations are speech protected by
47 USC 230.

This is a very hot and widely depated topic.

--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

The government defines what is not free speech by restricting it. If
you or your computer says XYZ and no one tries to stop you then its OK.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Jethro_uk" wrote in message news:u011g0$2lv5n$12@dont-email.me...

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without contravening
the 1st amendment because it was written by an AI engine.

Corollary: Would it be possible to ban a random string of numbers, as
they do not make up intelligible "speech" ?

And as a final kicker, could a state of affairs arise where AI output is >considered "not speech" (and therefore subject to bans) whilst
simultaneously being "speech" (and thus worthy of copyright and the
financial implications that brings) ?

If no one has asked these questions before, I find myself idly wondering
what a country with so many lawyers such as the US is doing with it's
time. Surely there aren't that many pictures of statues to occupy your >greatest minds ? (I write as a UKian watching from afar).

An AI is nothing more than a computer program. If a computer program has somehow been programmed to create speech or issue a document that includes libelous, pornographic or otherwise illegal material, then whoever was responsible for that computer program issuing that illegal material would be subject to the laws just as an actual human would.

--

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

In misc.legal.moderated, on Wed, 29 Mar 2023 07:19:59 -0700 (PDT),
Jethro_uk <jethro_uk@hotmailbin.com> wrote:

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without contravening
the 1st amendment because it was written by an AI engine.

I think the rights in the Constitution refer to the rights of people and
not machines. If someone would come forth as a proponent of whatever the
AI wrote, that would make it protected.

Corollary: Would it be possible to ban a random string of numbers, as
they do not make up intelligible "speech" ?

What would be the point of banning it, or distributing it? How would
one describe what is to be banned? I don't see it as likely anyone
would want to or that the government could ban something like that.

And as a final kicker, could a state of affairs arise where AI output is >considered "not speech" (and therefore subject to bans) whilst
simultaneously being "speech" (and thus worthy of copyright and the
financial implications that brings) ?

Books are being banned now, to a limited extent**, while still
retraining the copyrights their author has for them. **From school
libraries. Even at the worst and most ridiculous, they could still be at
the city public library, or at book stores, or for sale online. (Amazon
started out selling only books.) I've heard that 1984 has been banned
in some schools. Ironic, huh?

Also it would take some work by a lot of people, but one parent per
school or better yet, one child per school, could pass out lists of
what's been banned, both books and facts, where it can be read online,
what topics have been distorted, where the truth can be found online.
Online curricula could be posted that would include the parts of US
history that people in red states want to leave out.

A parent could stand at the street corner where children walk to school
and hand out one page that would give a central url where various topics
could be explained. The same URL could probably be used nationally
because "all liberals think alike"? Where most students take a school
bus, I can imagine the school wouldn't let the parent stand under the
school awning where the buses discharge their passengers, but his son or daughter could do that. At the entrance door in the morning or the kid
could do this at the cafeteria. He doesn't have to get every kid in one
day. When the kids learn about it they'll wait in line or get
hand-written copies of the url from their friends. Some won't but they
are probably the ones who would be the most immune to the information
even if the teacher delivered it.

Although the fact that the information is illicit in the view of the
school will probably make it 5 times as important in the view of the
students.

The news sometimes makes it sound like we'll have to wait a generation,
until Democrats are in control of a given city or state before children
can get a complete view of, for example, slavery (known as "involuntary relocation" according to some revisionists.) By the same token, those
same people or others might want to use this method to give another view
about abortion, which is so often described as health care, that it is
also fetus-killing.

I suppose in some schools they would try to stop the child from passing
out this mostly historical/scientific information sheet, but they would
lose. At least in public schools. Not sure about private schools and
wrt abortion maybe in parochial schools. But in private schools it's a different set of people deciding what will be taught from public
schools.

If no one has asked these questions before, I find myself idly wondering
what a country with so many lawyers such as the US is doing with it's
time. Surely there aren't that many pictures of statues to occupy your >greatest minds ? (I write as a UKian watching from afar).

Stautes? I don't know what you mean but that's probably because I'm not
one of my greatest minds.

--
I think you can tell, but just to be sure:
I am not a lawyer.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Josie Paine <cordon-invoice-photon@hash.fyi> wrote:

If the 1Ó½ æ- amendment is not extended to non-homo sapiens, let
alone non-Americans, why would it be extended to AI, at large? Its
developer could claim it's autonomously acting on behalf of
him/her. But a true, unowned autonomous AI would not have 1Ó½ æ-
amendment rights.

This reminds me of the case where a photographer set up a camera, and a
monkey ended up taking a shot that became very popular. The
photographer tried to get copyrights in the photo, but the court turned
him down, saying he didn't take it, and the monkey isn't entitled to copyrights.

--
Stu
http://DownToEarthLawyer.com


--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Rick" <rick@nospam.com> wrote:

"Jethro_uk" wrote in message
news:u011g0$2lv5n$12@dont-email.me...

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without
contravening the 1st amendment because it was written by an AI
engine.

Corollary: Would it be possible to ban a random string of numbers,
as they do not make up intelligible "speech" ?

And as a final kicker, could a state of affairs arise where AI
output is considered "not speech" (and therefore subject to bans)
whilst simultaneously being "speech" (and thus worthy of copyright
and the financial implications that brings) ?

If no one has asked these questions before, I find myself idly
wondering what a country with so many lawyers such as the US is
doing with it's time. Surely there aren't that many pictures of
statues to occupy your greatest minds ? (I write as a UKian
watching from afar).

An AI is nothing more than a computer program. If a computer
program has somehow been programmed to create speech or issue a
document that includes libelous, pornographic or otherwise illegal
material, then whoever was responsible for that computer program
issuing that illegal material would be subject to the laws just as
an actual human would.

Sure. But does the program/AI have rights that a human does to, say
copyrights or free speech? The answer normally is "no."


--
Stu
http://DownToEarthLawyer.com


--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Stuart O. Bronstein" wrote in message news:XnsAFD7AEF1CCEC5avocatstuyahoofr@130.133.4.11...

"Rick" <rick@nospam.com> wrote:

"Jethro_uk" wrote in message
news:u011g0$2lv5n$12@dont-email.me...

to be protected under the 1st amendment.

Or, to propose a reverse: Could an article be banned without
contravening the 1st amendment because it was written by an AI
engine.

Corollary: Would it be possible to ban a random string of numbers,
as they do not make up intelligible "speech" ?

And as a final kicker, could a state of affairs arise where AI
output is considered "not speech" (and therefore subject to bans)
whilst simultaneously being "speech" (and thus worthy of copyright
and the financial implications that brings) ?

If no one has asked these questions before, I find myself idly
wondering what a country with so many lawyers such as the US is
doing with it's time. Surely there aren't that many pictures of
statues to occupy your greatest minds ? (I write as a UKian
watching from afar).

An AI is nothing more than a computer program. If a computer
program has somehow been programmed to create speech or issue a
document that includes libelous, pornographic or otherwise illegal
material, then whoever was responsible for that computer program
issuing that illegal material would be subject to the laws just as
an actual human would.

Sure. But does the program/AI have rights that a human does to, say >copyrights or free speech? The answer normally is "no."

I don't think the AI per se has any "rights" because the AI is nothing more than a computer program. The real question may be can the output of a
computer program be subject to the laws governing free speech, copyright,
etc., and I suspect the answer in many cases is probably yes. If an AI program is allowed to create words that are a violation of law (again, libelous, pornographic, whatever), some human still has to arrange for that
AI output to be published or put in front of other people. Ultimately, it's the human (or human-controlled business) that allowed the output from the AI
to be published who is ultimately responsible.
--

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 31 Mar 2023 09:02:10 -0700, Rick wrote:

If an AI program is allowed to create words that are a violation of law (again, libelous, pornographic, whatever), some human still has to
arrange for that AI output to be published or put in front of other
people.

Not if it's published by an(other) algorithm, surely ?

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Jethro_uk" wrote in message news:u07eto$2lv5n$18@dont-email.me...

On Fri, 31 Mar 2023 09:02:10 -0700, Rick wrote:

If an AI program is allowed to create words that are a violation of law
(again, libelous, pornographic, whatever), some human still has to
arrange for that AI output to be published or put in front of other
people.

Not if it's published by an(other) algorithm, surely ?

Well again I have to go back to the fact that any AI is just a computer
program that is ultimately "hosted" somewhere. Some human or humans is in control of that program's execution and the dissemination of any output.
So if that AI's output is put before the public in any way - even if it is
put in front of another AI that can somehow do publishing or propagation - I still say there is ultimately a human or humans who control the first AI and where its output gets sent.

--

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 31 Mar 2023 16:27:32 -0700, Rick wrote:

"Jethro_uk" wrote in message news:u07eto$2lv5n$18@dont-email.me...

On Fri, 31 Mar 2023 09:02:10 -0700, Rick wrote:

If an AI program is allowed to create words that are a violation of
law
(again, libelous, pornographic, whatever), some human still has to
arrange for that AI output to be published or put in front of other
people.

Not if it's published by an(other) algorithm, surely ?

Well again I have to go back to the fact that any AI is just a computer program that is ultimately "hosted" somewhere. Some human or humans is
in control of that program's execution and the dissemination of any
output.
So if that AI's output is put before the public in any way - even if it
is put in front of another AI that can somehow do publishing or
propagation - I still say there is ultimately a human or humans who
control the first AI and where its output gets sent.

--

Sounds like we are accepting the Ghost in the Machine then ...

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Rick" <rick@nospam.com> wrote:

"Jethro_uk" wrote

Rick wrote:

If an AI program is allowed to create words that are a
violation of law
(again, libelous, pornographic, whatever), some human still has
to arrange for that AI output to be published or put in front of
other people.

Not if it's published by an(other) algorithm, surely ?

Well again I have to go back to the fact that any AI is just a
computer program that is ultimately "hosted" somewhere. Some
human or humans is in control of that program's execution and the dissemination of any output. So if that AI's output is put before
the public in any way - even if it is put in front of another AI
that can somehow do publishing or propagation - I still say there
is ultimately a human or humans who control the first AI and where
its output gets sent.

If the program is set up to write things at random, like an infinite
number of monkeys at an infinite number of typewriters, it would
eventually write Shakespeare, or defamatory material. If the program
writer can't get a copyright in the works of the AI (like the
photographer couldn't get copyrights in photos he set up to be taken by
a monkey), why should he be blamed for the defamation?

--
Stu
http://DownToEarthLawyer.com


--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"Stuart O. Bronstein" wrote in message news:XnsAFD95720396B2avocatstuyahoofr@130.133.4.11...

"Rick" <rick@nospam.com> wrote:

"Jethro_uk" wrote

Rick wrote:

If an AI program is allowed to create words that are a
violation of law
(again, libelous, pornographic, whatever), some human still has
to arrange for that AI output to be published or put in front of
other people.

Not if it's published by an(other) algorithm, surely ?

Well again I have to go back to the fact that any AI is just a
computer program that is ultimately "hosted" somewhere. Some
human or humans is in control of that program's execution and the
dissemination of any output. So if that AI's output is put before
the public in any way - even if it is put in front of another AI
that can somehow do publishing or propagation - I still say there
is ultimately a human or humans who control the first AI and where
its output gets sent.

If the program is set up to write things at random, like an infinite
number of monkeys at an infinite number of typewriters, it would
eventually write Shakespeare, or defamatory material. If the program
writer can't get a copyright in the works of the AI (like the
photographer couldn't get copyrights in photos he set up to be taken by
a monkey), why should he be blamed for the defamation?

Because he knowingly created (or at least enabled) the production of words
that could eventually be formed into material that could be against the law. Anyone knowingly creating/enabling an AI to produce words that can be seen
by the public would know there is this risk,

BTW, I'm pretty sure that if even all the monkeys in the world typed out
random letters continuously for as long as the universe has existed, they
would still have an extremely small chance of typing even a couple of
sentences from Shakespeare. The AI is clearly not just creating random
words but is following an algorithm created in a way that is designed to replicate human communication.

--

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

In misc.legal.moderated, on Sat, 1 Apr 2023 17:58:14 -0700 (PDT), "Rick" <rick@nospam.com> wrote:

If the program is set up to write things at random, like an infinite
number of monkeys at an infinite number of typewriters, it would ============================================================
eventually write Shakespeare, or defamatory material. If the program >>writer can't get a copyright in the works of the AI (like the
photographer couldn't get copyrights in photos he set up to be taken by
a monkey), why should he be blamed for the defamation?

Because he knowingly created (or at least enabled) the production of words >that could eventually be formed into material that could be against the law. >Anyone knowingly creating/enabling an AI to produce words that can be seen
by the public would know there is this risk,

BTW, I'm pretty sure that if even all the monkeys in the world typed out

You're moving the goal posts, or you're talking past Stuart. Stuart
referred to an infinite number of monkeys.

random letters continuously for as long as the universe has existed, they

It the number was infinite, I think one of them would type straight
through without any non-Shakespeare material. It woudln't take that
long, although supplying them all with paper and finding the one who
typed what you wanted would take a long time.

would still have an extremely small chance of typing even a couple of >sentences from Shakespeare.

You didn't, but some people make clear they think the monkey story is
about monkeys or how smart they are or typewriters, but it's not. It's
just to describe how big infinity is.

The AI is clearly not just creating random
words but is following an algorithm created in a way that is designed to >replicate human communication.

Good point.



--
I think you can tell, but just to be sure:
I am not a lawyer.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 02/04/23 04:52, Stuart O. Bronstein wrote:

"Rick" <rick@nospam.com> wrote:

"Jethro_uk" wrote

Rick wrote:

If an AI program is allowed to create words that are a
violation of law
(again, libelous, pornographic, whatever), some human still has
to arrange for that AI output to be published or put in front of
other people.

Not if it's published by an(other) algorithm, surely ?

Well again I have to go back to the fact that any AI is just a
computer program that is ultimately "hosted" somewhere. Some
human or humans is in control of that program's execution and the
dissemination of any output. So if that AI's output is put before
the public in any way - even if it is put in front of another AI
that can somehow do publishing or propagation - I still say there
is ultimately a human or humans who control the first AI and where
its output gets sent.

If the program is set up to write things at random, like an infinite
number of monkeys at an infinite number of typewriters, it would
eventually write Shakespeare, or defamatory material. If the program
writer can't get a copyright in the works of the AI (like the
photographer couldn't get copyrights in photos he set up to be taken by
a monkey), why should he be blamed for the defamation?

With that being said, a computer is not an autonomous, self-feeding system.
It needs to be executed, provided with electricity and paid for by someone.
You are not allowed to run a machine that with or without your knowledge,
by nature produces illegal combinations of letters. Why would we ever allow someone to run or execute something with no responsibility of such a device.
Or let a device exist that has no ownership or human to pin responsibility to?

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

In misc.legal.moderated, on Sun, 2 Apr 2023 22:23:50 -0700 (PDT), "Rick" <rick@nospam.com> wrote:

"micky" wrote in message news:n14i2idb775nhh61l45pdu1a85in9lrio4@4ax.com... >>

In misc.legal.moderated, on Sat, 1 Apr 2023 17:58:14 -0700 (PDT), "Rick" >><rick@nospam.com> wrote:

If the program is set up to write things at random, like an infinite >>>>number of monkeys at an infinite number of typewriters, it would >>============================================================
eventually write Shakespeare, or defamatory material. If the program >>>>writer can't get a copyright in the works of the AI (like the >>>>photographer couldn't get copyrights in photos he set up to be taken by >>>>a monkey), why should he be blamed for the defamation?

Because he knowingly created (or at least enabled) the production of words >>>that could eventually be formed into material that could be against the >>>law.
Anyone knowingly creating/enabling an AI to produce words that can be seen >>>by the public would know there is this risk,

BTW, I'm pretty sure that if even all the monkeys in the world typed out

You're moving the goal posts, or you're talking past Stuart. Stuart >>referred to an infinite number of monkeys.

I did move the goal posts, because talking about an infinite number of >anything is a purely abstract concept that doesn't really relate to anything >in the real world and certainly doesn't say much about AIs. But if we talk >about all the monkeys in the world or all the seconds that have gone by
since the Big Bang, those are at least finite concepts that we can kind of >relate to. My comment still stands that if all the monkeys in the world >(heck - I'll even say all the monkeys that have ever existed) - type random >characters for all the time since the Big Bang at say a character per
second, you still aren't likely to get much actual Shakespearian output.

Well, of course. That's the difference between 100 trillion monkeys and
an infinite number of monkeys. As I said, the story is about infinity,
not monkeys, no matter how big a finite number of them is, or how one
describes it.


--
I think you can tell, but just to be sure:
I am not a lawyer.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"micky" wrote in message news:n14i2idb775nhh61l45pdu1a85in9lrio4@4ax.com...

In misc.legal.moderated, on Sat, 1 Apr 2023 17:58:14 -0700 (PDT), "Rick" ><rick@nospam.com> wrote:

If the program is set up to write things at random, like an infinite >>>number of monkeys at an infinite number of typewriters, it would >============================================================
eventually write Shakespeare, or defamatory material. If the program >>>writer can't get a copyright in the works of the AI (like the >>>photographer couldn't get copyrights in photos he set up to be taken by
a monkey), why should he be blamed for the defamation?

Because he knowingly created (or at least enabled) the production of words >>that could eventually be formed into material that could be against the >>law.
Anyone knowingly creating/enabling an AI to produce words that can be seen >>by the public would know there is this risk,

BTW, I'm pretty sure that if even all the monkeys in the world typed out

You're moving the goal posts, or you're talking past Stuart. Stuart
referred to an infinite number of monkeys.

I did move the goal posts, because talking about an infinite number of
anything is a purely abstract concept that doesn't really relate to anything
in the real world and certainly doesn't say much about AIs. But if we talk about all the monkeys in the world or all the seconds that have gone by
since the Big Bang, those are at least finite concepts that we can kind of relate to. My comment still stands that if all the monkeys in the world
(heck - I'll even say all the monkeys that have ever existed) - type random characters for all the time since the Big Bang at say a character per
second, you still aren't likely to get much actual Shakespearian output.


--

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On March 31, Rick wrote:

I don't think the AI per se has any "rights" because the AI is nothing more than a computer program. The real question may be can the output of a computer program be subject to the laws governing free speech, copyright, etc., and I suspect the answer in many cases is probably yes. If an AI program is allowed to create words that are a violation of law (again, libelous, pornographic, whatever), some human still has to arrange for that AI output to be published or put in front of other people. Ultimately, it's the human (or human-controlled business) that allowed the output from the AI to be published who is ultimately responsible.

A person can hire an agent to act on his behalf. He is then liable
for the agent's actions. The AI program makes decisions autonomously,
on behalf of its owner, hence an agent.

--
Rich

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Wed, 24 May 2023 16:55:04 -0700, RichD wrote:

On March 31, Rick wrote:

I don't think the AI per se has any "rights" because the AI is nothing
more than a computer program. The real question may be can the output
of a computer program be subject to the laws governing free speech,
copyright, etc., and I suspect the answer in many cases is probably
yes. If an AI program is allowed to create words that are a violation
of law (again, libelous, pornographic, whatever), some human still has
to arrange for that AI output to be published or put in front of other
people. Ultimately, it's the human (or human-controlled business) that
allowed the output from the AI to be published who is ultimately
responsible.

A person can hire an agent to act on his behalf. He is then liable for
the agent's actions. The AI program makes decisions autonomously,
on behalf of its owner, hence an agent.

Does that suggest some form of vicarious liability ?

I would imagine a lot of gun (and ammunition) manufacturers are taking a
keen interest - to the extent of trying to influence - such developments.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)