4 Jun
2016
4 Jun
'16
6:21 p.m.
http://www.pcworld.com/article/3077977/data-center-cloud/googles-magenta-pr…
IMO, a lot of popular music is already machine-generated. ;)
i remember something along these lines from a few years back, an AI
project in the States that was far more advanced than Google's.
By using memory models, though, Google is making a good start. If
anyone can can find an article by George Rochberg called, "The
Avant-Garde and the Aesthetics of Survival," they will have an
excellent read.
One might observe that the machine wrote bad music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time. Still, i'm not
holding my breath, and it's certainly time to listen to that Bartok
concerto.
il lupo
4 Jun
4 Jun
9:21 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington wrote:
One might observe that the machine wrote bad music.
Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
Regards,
Ralf
10:01 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
For once I'm in total agreement with Ralf :o
Furthermore, there is not exactly a life-threatening shortage of music, so what
'need' does machine generated music fill?
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
Regards,
Ralf 
10:08 p.m.
New subject: Google Magenta project's first composition
On Sat, Jun 4, 2016 at 4:01 PM, Will Godfrey <willgodfrey(a)musically.me.uk>
wrote:
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
if this argument was totally solid, then there'd be no place for "process
music", which also has absolutely nothing to do with expressing emotions.
That means, for a start, no early Reichian minimalism. I don't want to live
in that world. Be my guest.
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
For once I'm in total agreement with Ralf :o
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
Regards,
Ralf 
10:18 p.m.
New subject: Google Magenta project's first composition
I would argue that this is "process music", and that the actual composers
are the people that programmed the AI.
Also: who says that Steve Reich has nothing to do with emotion? I find
that stuff to be beautiful.
On Sat, Jun 4, 2016 at 4:08 PM, Paul Davis <paul(a)linuxaudiosystems.com>
wrote:
On Sat, Jun 4, 2016 at 4:01 PM, Will Godfrey <willgodfrey(a)musically.me.uk>
wrote:
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
if this argument was totally solid, then there'd be no place for "process
music", which also has absolutely nothing to do with expressing emotions.
That means, for a start, no early Reichian minimalism. I don't want to live
in that world. Be my guest.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
For once I'm in total agreement with Ralf :o
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
Regards,
Ralf
10:22 p.m.
New subject: Google Magenta project's first composition
On Sat, Jun 4, 2016 at 4:18 PM, Robert Edge <thumbknucklerocks(a)gmail.com>
wrote:
I would argue that this is "process music",
and that the actual composers
are the people that programmed the AI.
That was precisely my point, mostly.
Also: who says that Steve Reich has nothing to do with emotion? I find
that stuff to be beautiful.
I also find it deeply beautiful. But Reich didn't compose it to convey a
state of mind (or heart).
10:49 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 16:22:38 -0400, Paul Davis wrote:
On Sat, Jun 4, 2016 at 4:18 PM, Robert Edge
<thumbknucklerocks(a)gmail.com> wrote:
Than this Google thingy is a tool, used for an opus, but not a machine
that does compose. This are two completely different approaches.
I would argue that this is "process
music", and that the actual
composers are the people that programmed the AI.
That was precisely my point, mostly. I also find it deeply beautiful. But Reich didn't
compose it to convey
a state of mind (or heart).
But he did, because he is a being with emotions. He wanted to do it,
because he has a live, experiences, passion and an impetus. A machine
has nothing of those qualities. If you program a machine as a tool,
used by you, it's not the same as programming a machine to do something
on it's own.
10:54 p.m.
New subject: Google Magenta project's first composition
On Sat, Jun 4, 2016 at 4:49 PM, Ralf Mardorf <ralf.mardorf(a)alice-dsl.net>
wrote:
But he did, because he is a being with emotions.
No, he didn't. Reich has stated very clearly the motivations and thoughts
that went into his early work. They had nothing to do with conveying
emotion.
11:03 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 16:54:01 -0400, Paul Davis wrote:
On Sat, Jun 4, 2016 at 4:49 PM, Ralf Mardorf
<ralf.mardorf(a)alice-dsl.net> wrote:
You completely don't understand what I explain. He decided not to play
soccer, he decided to make music.
n Sat, 4 Jun 2016 22:52:32 +0200, Robin Gareus wrote:
But he did, because he is a being with emotions.
No, he didn't. Reich has stated very clearly the motivations and
thoughts that went into his early work. They had nothing to do with
conveying emotion. On 06/04/2016 11:09 AM, tom haddington wrote:
Or making love/a baby computer while listening to music? It wouldn't
make sense, since computers are no living beings.
IIUC the coders intention wasn't to be the composers, by using this
Google thingy as a tool. The intention was to program a machine, that
has got the impetus itself.
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
Call me again once the google-cluster takes a break from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
Also only a matter of time?
Seen any long haired bit go byte-banging and bus-diving lately?! :)
11:10 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 23:02:55 +0200, Ralf Mardorf wrote:
On Sat, 4 Jun 2016 16:54:01 -0400, Paul Davis wrote:
More clearly:
The emotion is to decide to make music, instead of playing soccer. The
decision to make music without transporting an emotion, already is
transporting an emotion by the music, that shouldn't transport an
emotion.
So what's new with Google Magenta? Is it a machine that should be the
composer by itself? Or are the coders the composers? This are two
completely different approaches.
On Sat, Jun 4, 2016 at 4:49 PM, Ralf Mardorf
<ralf.mardorf(a)alice-dsl.net> wrote:
You completely don't understand what I explain. He decided not to play
soccer, he decided to make music.
But he did, because he is a being with emotions.
No, he didn't. Reich has stated very clearly the motivations and
thoughts that went into his early work. They had nothing to do with
conveying emotion. 
6 Jun
6 Jun
8:59 a.m.
New subject: Google Magenta project's first composition
On Sun, June 5, 2016 6:22 am, Paul Davis wrote:
On Sat, Jun 4, 2016 at 4:18 PM, Robert Edge
<thumbknucklerocks(a)gmail.com>
wrote:
Is/Are the parent/s responsible for the childs behaviour?
The parents of the AI are not necessarily able to control the output or
envision what the child will create. Does that mean they are composers,
conductors or engineers?
Is the audio engineer responsible for the quality of the performance?
--
Patrick Shirkey
Boost Hardware Ltd
I would argue that this is "process
music", and that the actual
composers
are the people that programmed the AI.
That was precisely my point, mostly.
4 Jun
4 Jun
10:44 p.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 16:08:14 -0400, Paul Davis wrote:
On Sat, Jun 4, 2016 at 4:01 PM, Will Godfrey
<willgodfrey(a)musically.me.uk> wrote:
That's wrong, "process music" underlies a human composer, the cognitive
process of a composer, the selection process of a composer. The emotion
not necessarily is what the music transports, the emotion could be,
what a composer animates to make music. It requires ardour (not the
software) to do this. A machine has got no impetus to make music. You
can use a machine for composing, you can use random processes, math
or what ever you want, as long as you chose this for an opus. If a
machine is intended not to be a tool, but also to fake impetus and
ardour, this is something completely different, but "process music".
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
if this argument was totally solid, then there'd be no place for
"process music", which also has absolutely nothing to do with
expressing emotions. That means, for a start, no early Reichian
minimalism. I don't want to live in that world. Be my guest. On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
For once I'm in total agreement with Ralf :o
One might observe that the machine wrote bad
music. Well, humans
are already doing that, too, so Magenta has gotten at least that
far! As with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine
generates is completely uninteresting to me, since a machine has
got no emotions I'm interested in. A machine has got no emotions
at all, so even if the machine would generate "good music", it
would be faked "good music", emotional fraud. Human impostors are
able to e.g. fake love. Victims often feel more loved by an
impostor, than by somebody who really loves them. Fraud could make
us feel good, we anyway dislike fraud. That just shows what kind
of company Google is. A human might be an untalented musician, but
at least a human usually has got real emotions. A machine that is
able to fake "good music" has got absolutely nothing to do with
progress. It's a damage. Developing something like this shows the
unimaginativeness of the developers. Nobody needs it, it's good
for absolutely nothing and even not a useful step to learn
something for useful AI projects or something like this.
Regards,
Ralf 
10:38 p.m.
New subject: Google Magenta project's first composition
On Sat, Jun 4, 2016 at 1:01 PM, Will Godfrey <willgodfrey(a)musically.me.uk>
wrote:
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
It's easy to argue that a machine has no soul, etc.
But, is machine produced music (or anything else for that matter) really
just machine produced music? Or is it music produced by a real human who
has directed or even created a machine to be an extension of himself?
Carrying the argument to it's silly end point, the only worthwhile music
would have to be vocals?
I might be able to create a machine to do work for me. But, it would never
produce (for example) rap music ... I'd just not add that capability.
--
**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob(a)mellowood.ca
WWW: http://www.mellowood.ca
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
For once I'm in total agreement with Ralf :o
Furthermore, there is not exactly a life-threatening shortage of music, so
what
'need' does machine generated music fill?
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
Regards,
Ralf 
5 Jun
5 Jun
7:40 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 13:38:03 -0700
Bob van der Poel <bob(a)mellowood.ca> wrote:
It's easy to argue that a machine has no soul,
etc.
But, is machine produced music (or anything else for
that matter)
really just machine produced music? Or is it music produced by a real
human who has directed or even created a machine to be an extension
of himself?
It will depend how much 'freedom' the machine has. Yes, persons built
it, but aren't those persons making their utmost effort to develop a
machine that can come up with decisions taken on its own, independent
from the creators ? Isn't that the premise of artificial intelligence ?
If not, we are also made, somehow. Biological machines. But we think we
are free.
As a spin to this, Elon Musk, of Telsa, SpaceX, Paypal, etc, declared
yesterday that "we are ‘almost definitely’ living in a Matrix-style
simulation"
7:43 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 21:01:16 +0100
Will Godfrey <willgodfrey(a)musically.me.uk> wrote:
Furthermore, there is not exactly a life-threatening
shortage of
music, so what 'need' does machine generated music fill?
Record companies not having to pay artists ? :)
9:28 a.m.
New subject: Google Magenta project's first composition
On 06/04/2016 07:43 PM, jonetsu(a)teksavvy.com wrote:
On Sat, 4 Jun 2016 21:01:16 +0100
Will Godfrey <willgodfrey(a)musically.me.uk> wrote:
> Furthermore, there is not exactly a life-threatening shortage of
> music, so what 'need' does machine generated music fill?
Presumably, the same 'need' that all other music fills. I recently read
a great book called "This Is Your Brain On Music" by Daniel J. Levitin
of the McGill University Laboratory for Musical Perception, Cognition
and Expertise. I heartily recommend reading it. Before going to college
and gaining his degrees, Dr Levitin was a session musician, sound
engineer and record producer for 10 years; worked with Stevie Wonder and
Blue Oyster Cult and a bunch of others.
The last chapter discusses the question: Why does music (and the unique
and powerful parts of the brain dedicated to music) exist? Until the
last 400-500 years, music was something everyone did, individually (to
demonstrate fitness as a mate) and as a group (to establish social and
group unity, enabling them to function together better when hunting or
facing danger), and also as a communication tool (think about the
Chinese language which uses pitch to differentiate words, or the many
varieties of whistles people have used to communicate across distances).
Music was also participatory: everyone sang, beat time, danced as part
of making music. (Dancing also demonstrated fitness as a mate as well as
your ability to work together as a group.) The modern classical concert
where everyone sits there engaged in their intellectual appreciation of
the violin soloist is completely alien to human music tradition.
Dr Levitin also points out that most people's emotional preference in
music is determined during adolescence, when the sex hormones start
firing madly and the body is screaming to its surroundings, "Reproduce
with me!" People fix on the music they think will get them laid.
Perhaps another need: The need of people to expand existing knowledge in
new ways. In Christian music, some have the concept of being "priests of
creation": People who enable nature to praise God in ways nature would
be otherwise unable to do on its own. Computers and software are part of
nature. So Magenta's developers are giving voice to otherwise voiceless
things like integrated circuits, silicon chips, etc.
Personally, I think that developers of audio creation and manipulation
software are also priests of creation. So bear your new title proudly!
Record companies not having to pay artists ? :)
Muzak?
Or record companies and artists both desperately trying to avoid having
some jury decide that a .2 second bit of sound on one musician's album
was really lifted from someone else's album? Or the chord progression in
your song was "stolen" from someone else? Such could be the idiocies of
our legal system, though.
--
David W. Jones
gnome(a)hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com
10:02 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 21:28:17 -1000
david <gnome(a)hawaii.rr.com> wrote:
Dr Levitin also points out that most people's
emotional preference in
music is determined during adolescence, when the sex hormones start
firing madly and the body is screaming to its surroundings,
"Reproduce with me!" People fix on the music they think will get them
laid.
There is a connection between sexuality and music, even with the much
reserved classical music. Eg. no need to bang drums to be sexual. But
is there also a more spiritual approach ? After all, when it comes
down to it, everything vibrates, every single piece of atom does. We
do not necessarily feel that, although isn't some music trying to make
a connection, a bridge, an experience, with "the universe" base don
frequencies ? For instance some Indian "om" mantras, buddhist gongs,
etc. This aspect, deep into the most superficial human, would evade
totally a machine.
Perhaps another need: The need of people to expand
existing knowledge
in new ways. In Christian music, some have the concept of being
"priests of creation": People who enable nature to praise God in ways
nature would be otherwise unable to do on its own.
It's interesting that the augmented 4th was termed the chord of evil
during the Renaissance and banned from church music. Today it is used
in emergency sirens. Some combinations of sounds do interact
"mechanically" (eg. without our conscious will) with our bodies. This,
though, could be put into the knowledge of a machine.
What makes humans tick can be programmed. Life itself cannot be known
by a machine. As some SF authors have written about (notably Frank
Herbert's "Destination: Void" original title "Do I wake or dream?"
in
1965) is the goal to make a machine that first, can learn by itself and
then, can eventually grasp the "meaning" of the universe ? Is the
Google Magenta a tiny step in that direction ?
7:11 a.m.
New subject: Google Magenta project's first composition
On 06/04/2016 12:21 PM, Ralf Mardorf wrote:
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
great emotional impact on audience does not require great emotional
investment of performer/author. Think of ocean or sunset or flowers - no
emotions but they are beautiful/impactful. (you might now change
argument to what's natural or not but that's a separate argument, only
pointing out the emotions here)
erik
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
7:33 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 22:12:24 -0700
Erik Steffl <erik(a)zasran.com> wrote:
great emotional impact on audience does not require
great
emotional investment of performer/author. Think of ocean or sunset or
flowers - no emotions but they are beautiful/impactful. (you might
now change argument to what's natural or not but that's a separate
argument, only pointing out the emotions here)
Even though humans are indeed machines in many aspects, as buttons can
be pressed to make them react in certain ways, emotionally, there is
still a chance that an emotional painting-by-numbers piece will not
fare so good, once the initial impact has passed.
As for the sunsets and such, there is more to the picture than meets
the eye.
8:41 p.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 01:33:43 -0400
"jonetsu(a)teksavvy.com" <jonetsu(a)teksavvy.com> wrote:
On Sat, 4 Jun 2016 22:12:24 -0700
Erik Steffl <erik(a)zasran.com> wrote:
How much of robots are we ? Can we assume that since we are humans, we
are not robots ? I'm thinking about that when working on a sketch (1).
The decision to press this button to play that beat then to press the
other button to segue into something else. The decision that yes, this
sounds all right with that. Apart from the guy who plays guitar out of
rhythm, anything else could have been done by a machine ?
(1) https://soundcloud.com/nominal6/jam5
great emotional impact on audience does not
require great
emotional investment of performer/author. Think of ocean or sunset
or flowers - no emotions but they are beautiful/impactful. (you
might now change argument to what's natural or not but that's a
separate argument, only pointing out the emotions here)
Even though humans are indeed machines in many aspects, as buttons can
be pressed to make them react in certain ways, emotionally, there is
still a chance that an emotional painting-by-numbers piece will not
fare so good, once the initial impact has passed.
8:52 p.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 14:41:14 -0400
"jonetsu(a)teksavvy.com" <jonetsu(a)teksavvy.com> wrote:
How much of robots are we ? Can we assume that since
we are humans,
we are not robots ? I'm thinking about that when working on a sketch
(1).
Speaking of robots, what are these 'entities' that sticks a 'I like' to
a song in soundcloud 1-2 minutes after it was uploaded ? For pete's
sake, the song lasts at least 3 minutes :)
6 Jun
6 Jun
9:16 a.m.
New subject: Google Magenta project's first composition
On Mon, June 6, 2016 4:41 am, jonetsu(a)teksavvy.com wrote:
On Sun, 5 Jun 2016 01:33:43 -0400
"jonetsu(a)teksavvy.com" <jonetsu(a)teksavvy.com> wrote:
At the moment there is a (subtle) interplay between the human engineers
and the AI. The AI can produce a vast array of results but the value is
currently at the discretion of the human engineer as they are the ones who
have to publish the results (and receive/absorb the criticism).
As the AI (child) learns though a results driven process it will
eventually become very competent at producing high quality and unique
results without the guidance of an engineer (parent).
How is that any different from the human learning process? Some of us are
faster/better than others but we all go through the same basic
machinations. Feedback is a very important contributor to behavior
(results).
--
Patrick Shirkey
Boost Hardware Ltd
On Sat, 4 Jun 2016 22:12:24 -0700
Erik Steffl <erik(a)zasran.com> wrote:
How much of robots are we ? Can we assume that since we are humans, we
are not robots ? I'm thinking about that when working on a sketch (1).
The decision to press this button to play that beat then to press the
other button to segue into something else. The decision that yes, this
sounds all right with that. Apart from the guy who plays guitar out of
rhythm, anything else could have been done by a machine ?
(1) https://soundcloud.com/nominal6/jam5
great emotional impact on audience does not
require great
emotional investment of performer/author. Think of ocean or sunset
or flowers - no emotions but they are beautiful/impactful. (you
might now change argument to what's natural or not but that's a
separate argument, only pointing out the emotions here)
Even though humans are indeed machines in many aspects, as buttons can
be pressed to make them react in certain ways, emotionally, there is
still a chance that an emotional painting-by-numbers piece will not
fare so good, once the initial impact has passed.
9:56 a.m.
New subject: Google Magenta project's first composition
On Mon, 6 Jun 2016 17:16:45 +1000 (EST), Patrick Shirkey wrote:
How is that any different from the human learning
process?
The answer quasi was given by Fons:
On Sun, 5 Jun 2016 09:50:49 +0000, Fons Adriaensen wrote:
* self-awareness: being aware of its own existence in
some
context, the way it is sustained physically, and being able
to act in ways that may affect this existence, and
* awareness of others entities who share the same or similar
conditions.
The context was a little bit different, but it remains true for
learning to do artwork, too. The computer has got absolutely no kind of
awareness, in the context of understanding life. It can not learn to
become creative.
AI can learn to distinguish words, but it even doesn't learn the
meaning of the words. The pattern detection learns by feedback to
distinguish words, but to understand what the words mean, it checks a
data bank and then follows conventionally programmed rules, to react.
It's similar for composing. Google pretend to do research regarding
artificial creativity, but the way they're doing it is ridiculous and
will lead to big nothing.
A human child learns to distinguish words, the meaning of the words, to
use the words creatively for life, in interaction with life experience,
since only this allows to learn the meaning and usage of the words.
AI as we know it has absolutely nothing in common with human intellect.
Regards,
Ralf
5 Jun
5 Jun
9:49 a.m.
New subject: Google Magenta project's first composition
On 06/04/2016 07:12 PM, Erik Steffl wrote:
On 06/04/2016 12:21 PM, Ralf Mardorf wrote:
Interestingly enough, color and music don't objectively exist. They are
both processed perceptions. Color is heavily processed, even before it
actually reaches the brain, by the optic nerve. Sound, on the other
hand, goes from the nerve cells (connected to hair cells) directly into
the brain, where parts of the brain analyze it into frequency components
while other parts analyze it into temporal components while other parts
analyze its timbre while other parts analyze the apparent spatial
location of the source, etc. Even our old reptilian brain is involved,
identifying emotional elements. (After all, one function of our
marvelous audio processing system in the past was to separate out the
single sound made by a tiger coming up behind you, and make you run like
hell before it could pounce!) All of this is kept in memory and analyzed
in real time by other parts of the brain as the brain builds an internal
model of the music, constantly predicts what will come next, and gets a
pleasure jolt every time its prediction proves correct. It also gets a
pleasure jolt when its prediction is wrong but the brain successfully
reworks its model of the music to achieve another correct prediction.
(Damn, music processing is one of the most complex things brains do!)
All done so that we can take in a barrage of sound and separate it into
the bass playing this, the keyboard playing that, the rhythm guitar
strumming this way, the singer singing a line and the lead guitar
playing a short riff at the end of each line - while being aware from
how she sounds that the hot chick beside you at the concert is really
getting turned on ...
--
David W. Jones
gnome(a)hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com
On Sat, 4 Jun 2016 05:09:02 -0400, tom haddington
wrote:
great emotional impact on audience does not require great emotional
investment of performer/author. Think of ocean or sunset or flowers - no
emotions but they are beautiful/impactful. (you might now change
argument to what's natural or not but that's a separate argument, only
pointing out the emotions here) One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
The point isn't, if a machine is able to fake music, it doesn't
matter, if it's good or bad faked music. What the machine generates is
completely uninteresting to me, since a machine has got no
emotions I'm interested in. A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked "good
music", emotional fraud. Human impostors are able to e.g. fake love.
Victims often feel more loved by an impostor, than by somebody who
really loves them. Fraud could make us feel good, we anyway
dislike fraud. That just shows what kind of company Google is. A human
might be an untalented musician, but at least a human usually has got
real emotions. A machine that is able to fake "good music" has got
absolutely nothing to do with progress. It's a damage. Developing
something like this shows the unimaginativeness of the developers.
Nobody needs it, it's good for absolutely nothing and even not a
useful step to learn something for useful AI projects or something
like this.
7:47 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked
"good music", emotional fraud.
Google Magenta has read this and it makes it sad.
<oo>
8:46 a.m.
New subject: Google Magenta project's first composition
On 06/04/2016 07:47 PM, jonetsu(a)teksavvy.com wrote:
On Sat, 4 Jun 2016 21:21:10 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
FWIW, I think a true artificial intelligence would have
feelings/emotions, or something equivalent. And ones based on learning
could well learn those feelings from whatever they're learning from. In
Magenta's case, that would be the human music that it's been learning from.
Way back in the days of DOS, I remember a poetry-writing program that
did a fair job of writing basic modern poetry. Occasionally it was even
witty. It wasn't even close to an intelligence; it was driven by
stochastic rules the programmer had built into it, couldn't change the
rules, didn't take in any feedback, or even read its own output.
I think Google's Magenta is more like what you might get if you applied
Bayesian statistical techniques based on undirected self-analysis of
existing music, defined a "goal" for the system to achieve, and added
whatever generated music the system came up with as feedback into the
process. Perhaps along with feedback from entities outside the system
(humans). (As if I know anything about that.)
Anyway, music and emotions aren't necessary connected. I'm sure you've
all heard musicians who have remarkable technical mastery, but whose
playing can only be described as emotionally cold. I think you've
probably heard compositions, too, that could be described best as
intellectual exercises in making complex patterns, and also lack
emotional feeling.
--
David W. Jones
gnome(a)hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com
A machine has got no emotions at all, so
even if the machine would generate "good music", it would be faked
"good music", emotional fraud.
Google Magenta has read this and it makes it sad.
<oo>
9:19 a.m.
New subject: Google Magenta project's first composition
On Sat, 4 Jun 2016 22:12:24 -0700, Erik Steffl wrote:
great emotional impact on audience does not require
great emotional
investment of performer/author.
To decide to make music, because you like this kind of art is emotional.
Why don't we play soccer, run amok, plant a tree at the time we decide
to make music? We decide to make music, experienced a sunset, by love,
by hate, because we are living. As Robin already pointed out, does a
computer enjoy listening to music? If not, why does it compose music?
Some humans make "music" just to make money and even those have an
impetus, a machine doesn't have. I even couldn't stand to read the
full original website of the Magenta project. Actually they don't know
what they are doing and they hope it will be a tool for artists. For
what purpose? They try to make a tool without a real concept and
hope somebody could find it useful. That's a damage. It even wouldn't be
a tool for "process music" assumed it automatically composes, while the
impetus was a composer turning on the machine. They try to find out
what the impetus of human artists is and to make it part of the
software, of what they guess AI is. Impetus isn't part of intelligence,
it's part of living beings independent of intelligence. I've got the
impression they are living in a madhouse and/or are 2 years old.
Joseph Weizenbaum (8 January 1923 – 5 March 2008) was and still is one
of my idols and one of the few with knowledge about AI, since he's one
of the fathers of AI and at the same time he was a critics, a sceptic
for good reasons.
While there is something we call AI, we are able to fake human
intelligence even without using AI at all. I'm an atheist, humans are
some kind of machine, too, but we are alive. To make art being alive
is a basic requirement.
On Sat, 4 Jun 2016 22:16:53 -0700, Erik Steffl wrote:
On 06/04/2016 01:52 PM, Robin Gareus wrote:
Joy is a matter of simplicity, not complexity. Actually it would
require "artificial" live, not just a simulation of artificial live,
let alone a simulation of human intelligence, is in any way useful to
understand joy, impetus, emotions.
On Sun, 5 Jun 2016 01:33:43 -0400, jonetsu(a)teksavvy.com wrote:
Call me again once the google-cluster takes a
break from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
Also only a matter of time?
Seen any long haired bit go byte-banging and bus-diving lately?! :)
matter of complexity. One can say matter of time as a shorthand for
"matter of complexity and complexity will increase with time". Even though humans are indeed machines in many aspects,
as buttons can
be pressed to make them react in certain ways, emotionally, there is
still a chance that an emotional painting-by-numbers piece will not
fare so good, once the initial impact has passed.
As for the sunsets and such, there is more to the picture than meets
the eye.
That's the problem with those Google folks. They believe in biased
philosophy, biased neuroscience and started to program. It's even not
a well-wrought experiment.
On Sun, 5 Jun 2016 01:40:12 -0400, jonetsu(a)teksavvy.com wrote:
It will depend how much 'freedom' the machine
has. Yes, persons built
it, but aren't those persons making their utmost effort to develop a
machine that can come up with decisions taken on its own, independent
from the creators ? Isn't that the premise of artificial
intelligence ?
There are different approaches for what AI is used for and what AI
means.
If not, we are also made, somehow. Biological
machines. But we think
we are free.
Apart from what ever neuroscience claims, our free will at least is
limited by our abilities, e.g. the acoustic range, by our life
experience. However, we have individual meanings of life, at least an
attitude regarding a meaning of life. A computer has got no reason to
compose, the machine needs patterns, an acoustic range, a definition
of what music is, rules of existing music styles. The coders are
the composers, this is nothing new and has nothing to do with
intelligence at all. Another target is to give the computer life, make
it alive, give it the impetus for self-preservation etc.. This requires
a biological computer. Introducing new music styles, new "rules" and
even using existing ones for joy, by passion, requires to be alive.
Music is about life, it is about emotions, even if a composer has a
thought about not transporting emotions, this aspect still is part of a
composition.
Regards,
Ralf
9:33 a.m.
New subject: Google Magenta project's first composition
On 06/05/2016 12:19 AM, Ralf Mardorf wrote:
On Sat, 4 Jun 2016 22:12:24 -0700, Erik Steffl wrote:
why did you remove the part when I write about how the author might
not be human or intelligent in any sense at all (flower, sunset etc.).
That was the point.
...
great emotional impact on audience does not
require great emotional
investment of performer/author.
To decide to make music, because you like this kind of art is emotional.
Why don't we play soccer, run amok, plant a tree at the time we decide
to make music? We decide to make music, experienced a sunset, by love, On Sat, 4 Jun 2016 22:16:53 -0700, Erik Steffl wrote:
emotions only come up in systems that are complex enough (that
doesn't mean that all complex systems would have emotions, just that the
current AI is too simple to have emotions and that complex enough AI
*might* eventually have emotions).
erik
On 06/04/2016 01:52 PM, Robin Gareus wrote:
Joy is a matter of simplicity, not complexity. Actually it would
require "artificial" live, not just a simulation of artificial live,
let alone a simulation of human intelligence, is in any way useful to
understand joy, impetus, emotions. Call me again once the google-cluster takes a
break from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
Also only a matter of time?
Seen any long haired bit go byte-banging and bus-diving lately?! :)
matter of complexity. One can say matter of time as a shorthand for
"matter of complexity and complexity will increase with time".
9:57 a.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 00:33:27 -0700, Erik Steffl wrote:
Emotions are required, they fulfil a "function" of life. they are the
cause to protect ourself, to be kindhearted, human etc., nothing like
that is required for a machine to exist. The sunset itself doesn't make
music, even the flower, a living being doesn't make music.
On 06/05/2016 12:19 AM, Ralf Mardorf wrote:
The sunset happens without intelligence or emotion at all and there's
no doubt that it could cause emotions to somebody watching the sunset.
On Sat, 4 Jun 2016 22:12:24 -0700, Erik Steffl
wrote:
why did you remove the part when I write about how the author might
not be human or intelligent in any sense at all (flower, sunset etc.).
That was the point. great emotional impact on audience does not
require great emotional
investment of performer/author.
To decide to make music, because you like this kind of art is
emotional. Why don't we play soccer, run amok, plant a tree at the
time we decide to make music? We decide to make music, experienced a
sunset, by love, On Sat, 4 Jun
2016 22:16:53 -0700, Erik Steffl wrote:
emotions only come up in systems that are
complex enough (that
doesn't mean that all complex systems would have emotions, just that
the current AI is too simple to have emotions and that complex enough
AI *might* eventually have emotions).
10:17 a.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 09:19:26 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
As Robin already
pointed out, does a computer enjoy listening to music? If not, why
does it compose music?
Iconoclast Kaiji Haino once said that sound needs air. That it is not
possible to embody sounds, music, if there is no air. Considering
this, then how would music composed with no air, composed by an entity
that cannot listen to it through air, fare ?
The beginning of this lecture expresses the need to make the sound
travel, to not control it so much (lecture starts after he stops
banging on some discs. And despite that yes, the snob factor is quite
high, the ideas he expresses at the beginning complements this
discussion on computer-generated music:
https://www.youtube.com/watch?v=xyIEbcwz_n8
10:54 a.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 04:17:19 -0400, jonetsu(a)teksavvy.com wrote:
On Sun, 5 Jun 2016 09:19:26 +0200 Ralf Mardorf wrote:
Actually the machine could listen by microphone, OTOH the computer
even could consider gravity waves or light as being sound. By nature
the computer isn't limited to ears. A computer even could hear by
seeing and a computer doesn't need to perform a mating dance. We can
not hear with our eyes. Human synaesthesia is not the same. A computer
has got not the limits a human has got, OTOH even deaf people make
music, they e.g. feel vibrations, there's a famous deaf female
percussionist, I'm not thinking about composers who lost hearing.
http://www.google.de/search?q=deaf+female+percussionist&hl=en
The biggest difference between machines, even "intelligent" machines of
the future and a living musician is, life and the sense of music. No
sex, no music.
Likely somebody will now mention an asexual composer of "process
music" and maybe the composer in addition suffers from a personality
order or brain damage, so that the composer has got absolutely no
emotions. In addition this composer is the only person with a kind of
synaesthesia, that makes the composer really hear light. This composer
grows up in a small jail, with less stimuli.
This composer might exist, but IIUC this is not what Google tries to
develop with Magenta. Such a composer unlikely is heard by many people,
even "process music" from composers who are well known for their
emotional conventional music, isn't that much heard by a larger
audience.
As Robin already pointed out, does a computer
enjoy listening to
music? If not, why does it compose music?
Iconoclast Kaiji Haino once said that sound needs air. That it is not
possible to embody sounds, music, if there is no air. Considering
this, then how would music composed with no air, composed by an entity
that cannot listen to it through air, fare ?
8:56 p.m.
New subject: Google Magenta project's first composition
On Sun, Jun 5, 2016 at 1:54 AM, Ralf Mardorf <ralf.mardorf(a)alice-dsl.net>
wrote:
The biggest difference between machines, even "intelligent" machines of
the future and a living musician is, life and the sense of music. No
sex, no music.
Ahh, now I know why my saxophone playing is so shitty :)
--
**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob(a)mellowood.ca
WWW: http://www.mellowood.ca
10:01 p.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 11:55:57 -0700, Bob van der Poel wrote:
On Sun, Jun 5, 2016 at 1:54 AM, Ralf Mardorf wrote:
Siri recommends objectophilia, a candlelight dinner with soft music from
Magenta and Paro http://www.parorobots.com/ will fall in love with you.
No sex, no music.
Ahh, now I know why my
saxophone playing is so shitty :)
10:07 p.m.
New subject: Google Magenta project's first composition
On Sun, 5 Jun 2016 22:01:03 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
Siri recommends objectophilia, a candlelight dinner
with soft music
from Magenta and Paro http://www.parorobots.com/ will fall in love
"Robotic seals comfort dementia patients..."
Maybe the human race do need robots after all.
10:20 p.m.
New subject: Google Magenta project's first composition
On Sun, Jun 5, 2016 at 1:07 PM, jonetsu(a)teksavvy.com <jonetsu(a)teksavvy.com>
wrote:
On Sun, 5 Jun 2016 22:01:03 +0200
Ralf Mardorf <ralf.mardorf(a)alice-dsl.net> wrote:
Might be just in time to replace us all. And, they might do a better job
looking after our planet. I say "let's give them a chance!".
--
**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob(a)mellowood.ca
WWW: http://www.mellowood.ca
Siri recommends objectophilia, a candlelight
dinner with soft music
from Magenta and Paro http://www.parorobots.com/ will fall in love
"Robotic seals comfort dementia patients..."
Maybe the human race do need robots after all. 
4 Jun
4 Jun
10:52 p.m.
New subject: Google Magenta project's first composition
On 06/04/2016 11:09 AM, tom haddington wrote:
One might observe that the machine wrote bad music.
Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
Call me again once the google-cluster takes a break from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
Also only a matter of time?
Seen any long haired bit go byte-banging and bus-diving lately?! :)
ciao,
robin
5 Jun
5 Jun
7:16 a.m.
New subject: Google Magenta project's first composition
On 06/04/2016 01:52 PM, Robin Gareus wrote:
On 06/04/2016 11:09 AM, tom haddington wrote:
matter of complexity. One can say matter of time as a shorthand for
"matter of complexity and complexity will increase with time".
erik
One might observe that the machine wrote bad
music. Well, humans are
already doing that, too, so Magenta has gotten at least that far! As
with chess machines, it may be a matter of time.
Call me again once the google-cluster takes a break from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
Also only a matter of time?
Seen any long haired bit go byte-banging and bus-diving lately?! :) 
11:50 a.m.
New subject: Google Magenta project's first composition
On Sat, Jun 04, 2016 at 10:52:32PM +0200, Robin Gareus wrote:
Call me again once the google-cluster takes a break
from indexing and
processing in order to attend a concert and listen to some music for
pure enjoyment.
To the point, I'd say.
For me anything called 'art' becomes interesting (and then
I'd call it 'art' - but that's just a convenient definition
of a word) if to the observer it refers in some way - maybe
very indirectly - to the experience of being a human.
In that sense there's no doubt that e.g. Reich's music is
'art'. Even if he didn't want to convey any feeling or
emotion, just the simple fact of knowing that one of us
had the idea of having microphones swing on a pendulum
above a speaker and let them produce sound by feedback,
produces an 'emotion ' by revealing something we're
capable of.
AI as we know it isn't anywhere near to being able to produce
anything like that by itself. It would requires at least two
steps:
* self-awareness: being aware of its own existence in some
context, the way it is sustained physically, and being able
to act in ways that may affect this existence, and
* awareness of others entities who share the same or similar
conditions.
The latter can be tricky even for us humans. We can both overdo
it (by assuming human feelings in things that aren't capable
if having them), and fail miserably (e.g. us vs. them tribal
thinking, racism).
On a different note: whatever Google produces isn't about
art. Google's business is selling surveillance data. If you
use Google's services, you're not a user, let a alone a
customer, you're the product being sold. Better be aware
of that.
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
2891
days inactive
2893
days old
linux-audio-user@lists.linuxaudio.org
37 comments
12 participants
participants (12)
-
Bob van der Poel -
david -
Erik Steffl -
Fons Adriaensen -
jonetsu＠teksavvy.com -
Patrick Shirkey -
Paul Davis -
Ralf Mardorf -
Robert Edge -
Robin Gareus -
tom haddington -
Will Godfrey