Cheating

I do not want to sound offensive in any way shape or form towards David
Cope, and indeed, I found his books on EXPERIMENTS ON MUSICAL INTELLIGENCE,
about the exploits on SARA so interesting, that I went and brought both
books, the sequel, and the origional, however, I have one grudge about SARA.

This algorithm,-It is, what, -correct me if I'm wrong, they call
"parasitic", parasitic in the sence that it used fragments of music
precomposed by a human and then mixes them up.

I would only give full credit to a computer programme that started "RIGHT
FROM SCRATCH", and used an algorithm that made beautiful music from purely
random numbers, that were generated, from whatever source we would have,
ideally radioactive decay, but that is another story.

I mean, I could say:"Hooray, I've written an AI programme that generates
poetry....", and lead you in the direction of a word processor, a word
processor to which I generated most, if not nearly all of the input.

So, how do we define a TRUELY artificially intelligent algorithm for making
music, music that would fool an audience?

Is this a valid argument?

A sound argument?

You're thoughts please.

Well ...

First, let me invite you to join the "algo-comp" group at eGroups ... we
discuss these issues at great length. Now, into the breach :-)

> I do not want to sound offensive in any way shape or form towards David
> Cope, and indeed, I found his books on EXPERIMENTS ON MUSICAL
> INTELLIGENCE,
> about the exploits on SARA so interesting, that I went and brought both
> books, the sequel, and the origional, however, I have one grudge
> about SARA.
>
> This algorithm,-It is, what, -correct me if I'm wrong, they call
> "parasitic", parasitic in the sence that it used fragments of music
> precomposed by a human and then mixes them up.

Well, EMI and SARA are a bit more complex than that. You have the books, as
do I. If you're not a LISP aficionado, you'll find them difficult reading,
since much of the important detail is in the code. Also, if you don't have a
Macintosh, the CD that comes with the more recent book cannot be read. A
pity, since I have two LISP environments for my Windows box and would love
to port this code and play with it.

[snip]

> So, how do we define a TRUELY artificially intelligent algorithm
> for making
> music, music that would fool an audience?
>
> Is this a valid argument?
>
> A sound argument?
>
> You're thoughts please.

Well ... I've heard the music produced by Cope's scheme. It is pleasant
enough, but it just doesn't have the spark that Bach, Mozart, Beethoven,
Chopin, Scott Joplin and Stravinsky have. It's a better imitation of them
than, say, Illiac Suite is of Palestrina, but it still is an imitation.

Would it fool an audience? Probably not, even people who didn't know all of
Beethoven's sonatas, for example.
--
M. Edward Borasky, Borasky Research
http://www.borasky-research.com
mailto: znmeb@borasky-research.com

"There's no fuel like an old fuel" -- the National Coal Institute

Why start with random numbers? Do you think human composers do, apart from people such as John Cage? Isn't that a bit like expecting monkeys at typewriters to turn out the works of Shakespeare, eventually? And why radioactive decay as a source of random numbers? Is that more likely to create "le jazz hot"?

>>> thcdelta@ihug.co.nz 01/15/01 09:11PM >>>
I do not want to sound offensive in any way shape or form towards David
Cope, and indeed, I found his books on EXPERIMENTS ON MUSICAL INTELLIGENCE,
about the exploits on SARA so interesting, that I went and brought both
books, the sequel, and the original, however, I have one grudge about SARA.

This algorithm,-It is, what, -correct me if I'm wrong, they call
"parasitic", parasitic in the sense that it used fragments of music
precomposed by a human and then mixes them up.

I would only give full credit to a computer program that started "RIGHT
FROM SCRATCH", and used an algorithm that made beautiful music from purely
random numbers, that were generated, from whatever source we would have,
ideally radioactive decay, but that is another story.

I mean, I could say:"Hooray, I've written an AI program that generates
poetry....", and lead you in the direction of a word processor, a word
processor to which I generated most, if not nearly all of the input.

So, how do we define a TRULY artificially intelligent algorithm for making
music, music that would fool an audience?

Is this a valid argument?

A sound argument?

You're thoughts please.

You do not need web access to participate. You may subscribe through
email. Send an empty email to one of these addresses:
tuning-subscribe@egroups.com - join the tuning group.
tuning-unsubscribe@egroups.com - unsubscribe from the tuning group.
tuning-nomail@egroups.com - put your email message delivery on hold for the tuning group.
tuning-digest@egroups.com - change your subscription to daily digest mode.
tuning-normal@egroups.com - change your subscription to individual emails.