Markov or stochastic music had been around for a long time before Tech Fantasies had started using it to generate audible school and student profiles. There was enough published research and open-source software available that Ann, Drake, Sally and their many colleagues had easily made a first cut at the desired software. Even that rather crude approximation had astonished people, but Kelly and her friends were not satisfied.
“It’s not good enough, Mommy”, Kelly said. “Suzy and I have being playing with it a lot. It works, but is just not good enough.” Kelly liked all kinds of music. Her Level Five best friend, Suzy Nicholson, liked mostly classical music.
“I’ve applied it to my own playlists, which are all classical, Mrs. Aston”,
Suzy said. “The music that comes out is not very sophisticated.”
“It’s not all that great for other kinds of music, Mommy. Being able to generate new music from collections of music is wonderful, and it does let us compare profiles, but the generated music itself is not magical stuff.”
“I know, Kelly. None of us knew much about music. We tried to solve the technical problems. There were plenty of those. More than enough. Control of repetition was the worse.”
“How is that, Mrs. Aston”, Suzy asked.
“Well, music is repetitive by nature, but that poses problems for Markov processes in which what comes later is based on what came before. We had to do all sorts of weighting tricks to keep it from being too repetitive — just playing the same thing over and over. We had to take the music apart, too, so that we did separate Markov processes on rhythms, chords and melodies. That was the hard part. None of us knew enough about music to do that analysis well. That’s probably the real reason why you aren’t satisfied with the result.”
“You must have help, though? People who do know music, don’t you?”, Kelly asked her mother.
“Yes, we had some help. We wouldn’t have been able to do it at all otherwise. We made use of some existing software, too. That made up for some of what we didn’t know ourselves, though using other people’s software can sometimes be harder than writing your own. I know it’s not perfect, but it is more than just a proof of concept. This has become a great project.”
“I guess there is nothing we could do to help is there? Maybe Suzy could. She knows everything about music.”
Sally didn’t want to cut Kelly out of the loop. Her lovely daughter was her life. Besides, Kelly had almost invented the project, though Drake had the key vision.
“Maybe you could both help”, Sally told the girls. “If I were to package some of our components so they could be called from the school’s ST language, maybe you could play around with them. Some of your other friends might be interested, too. If people at the school worked on it enough, they might be able to take it over, make it their own. What we have is too big and complicated, it would be nice to have something simpler, which students could understand and maintain themselves.”
“Well, I do know something about music and can program a little bit in ST”, Kelly said. “I don’t think Suzy can program at all, but she knows everything about music, right Suze?”
“No, but some. I like to listen to classical music, which is mostly what I play, but — “
“ — but she can play anything. I’ve heard her. Jazz, country, pop, anything”, Kelly finished.
“Don’t worry too much about the programming, Kelly. There are other students who can help with that. Though you might have fun trying to write a tiny Markov music program in ST. I’ll give you some information about how one works. Then if you want to, you might just start from scratch, see what you get. What I’d really like to see though, is more analysis of how music fits together.”
“We can find our for ourselves how one works”, Kelly said. “I am sure it is on the Internet somewhere, and Suzy can handle the music stuff.”
“I am sure all the information you need is out there. Try searching for Markov music or stochastic music.”
Enthusiastic, Suzy and Kelly went off to try to create a toy version of the new software. “This is just for fun, Suze. And to learn something. I’ll think about what other kids might do, if we can get more.”
The current Markov music software was being constantly improved, but could still be put in package form. It would take a while for Sally and her colleagues to collect the components and put them in easily callable form, but it would be worth doing even if nobody at the school ever tried to use them.
Everyone at school did use the current music generation program. Mr. Everett collected the playlists from everyone in his homeroom class of moderately compatible Grade Seven students. They then spent every morning for a week listening to the generated music, which more or less defined their class.
Kelly could not just sit still and listen to music. The made notes on paper, drew diagrams and typed things into her computer while the music played.
At the end of Friday afternoon, Kelly found Mr. Everett and handed him a printout.
“This is a project, sir. I might do it with Suzy, if nobody else is interested. Please tell me what you think.”
“OK, I’ll read this, Kelly. But can you give it too me in a few words, right now?”
“I think what my mother would most like to see is a complete revision of the music generator program, written entirely in ST. The trivial one was easy. It is just an ST program that asks you if you want to run the main program. I also wrote the next step, which is an ST program which calls the main program’s playlist input function, then the main program’s music generation function. I thought we could keep splitting it up into smaller and smaller pieces, until we got to simple functions. I don’t know what to do after that, though. Get help, I guess.”
“That’s good, Kelly. I’ll read what you have written here, then perhaps talk to the class about it Monday morning.”
As use of the existing software spread its power was better understood, but so were its limitations. The impact of music seemed to be on the right brain and also on the limbic system, which was a locus for emotions. The left brain cognitive functions were not so well represented or responsive.
The finest application of the stochastic music generation system was with roommates in the school’s residences. Two very close friends at Level Five or above could share a room under certain conditions. Couples with a sexual relationship could only do so if of legal age in both the state and in their home countries. Students with homes in the city or a plausible commuting distances were excluded from residences need for others.
Wherever students have shared rooms, music has always been a problem because tastes do differ. One person’s music is another’s noise. The new music generation system could guarantee the creation of music both roommates would like. Each of the two people sharing a room would have a playlist, specifying a set of pieces the person liked to hear. The intersection of those two sets was music they both liked to hear. From that intersection could be generated an endless supply of music which both roommates would like or at least find inoffensive. That could be good background music for when they wanted some, but was not the most interesting use of their playlists. Another approach would be to take the union of those sets, instead of their intersection. This music would usually be richer, drawing on more sources, but not necessarily something either would like very much. For example, some people do not consider rap music to be music and want to turn it off when it is played. Using the union of one of their playlists with one from someone who liked rap music a lot would produce a mixture that at least one of them would probably find objectionable.
This is where the unique circumstances of the school residences became manifest. Given that the two individuals were already known to be Level Five or above, their musical preferences were unlikely to be that different. When two people have similar tastes in music it is usually easy for each to learn or acquire a taste for what the other likes.
The twelve year old boy Santos Coriega in Kelly’s homeroom was from Mexico City and liked Latin music. He shared a room with a boy from Louisiana who liked fusion music with an element of New Orleans jazz in it. The intersection of their playlists was mostly pop, though pop using brass section seemed to appeal to both. The union of their playlists was very large. The music created from them was very rich. It took a while, but soon both boys liked it much better than the music they had grown up with.
A true understanding of what this meant began in the residences, and it was Santos who explained it to Kelly. “What Rene and I did, Kelly, be merged our playlists, just the two of us. The result was fantabulous because we are so the best of amigos. Lots better stuff than we got with more people. Hey, I thought, wow, this is great, the two of us. Then it hit me. The whole school, all of us, we chain. Then wham, everybody in the house knows.”
“Right. Each one of us has someone really close. What we get from merging just our two playlists is what two Level Five people have. Compatible mean right and left brains, so if two compatible people do it, the music is not just right brain stuff.”
“What I said.”
Kelly and Suzy had each heard the rich music that the school generated. Each girl had turned her own playlist into an endless stream of music. Fascinated, neither was satisfied. They had collected a few playlists from various friends and merged them. Good, but not yet spectacular. One simple thing that both girls had overlooked was the most obvious, merging just their own two playlists, nobody else’s.
The result was not so immediately magical, not so nearly visceral. It took time to fully appreciate. It was not as rich as music based on a larger number of source playlists. But it was more profound. It was the music of two rather musical Level Five friends. Drawing from and reflecting that level of compatibility, it captured something from all parts of their brains, not just those traditionally associated with music.
Kelly had one other Level Five friend, Cariola Tevanza. Cariola was interested in math and computer science, like Kelly’s parents.
“Cariola. We have to try this. We have to try this. Merge our playlists, just the two of us. Nobody else.”
“I thought it was better with more. It always sounds better to me, though not really the kind of music I like.”
“I know, I know. But we are real compatible. That makes a difference. It does.”
Cariola liked choral music, from any time period, or in any style, just as long as it was sung by a choir. The union or merger of their playlists was very eclectic, with more music for multiple voices, but far beyond any superficial differences was the overwhelming effect of unity which the music generated in both girls. It made them feel an intellectual couple. They could sense the linkings of their minds.
Kelly went to see the school principal herself. “Dr. Grey?”
“Yes, Kelly. This is a surprise. What’s up?”
“We need to get every compatible pair to merge their playlists, sir. Ever person has got to do it. Just in pairs. If you have two friends, do it separately with each one. Try it yourself, sir!”
“You are sure about this?”
“I wouldn’t bother you if I wasn’t, Dr. Grey.”
So Paul Grey contacted his three close friends, two students and a teacher. Individually he merged his personal playlist with each of their. In pairs, they sat back to listen.
Soon as excited as Kelly was, Paul Grey broadcast an e-mail message to every person in the school. He broadcast it also to school members who did not actually attend the school, such as important people in the local Tech Fantasies organization. He included long-distance links. This would be overkill, of course. Everybody might already know. Information passed through the school like magic with no electronic help.
By Monday morning, when Mr. Everett spoke to his class about Kelly’s project, the whole class had tried pair-music generation with all their compatible friends. Amazed at the results, they were receptive to anything to do with music generation. After giving a bit of an explanation, Mr. Everett said, “This would make a nice class project, and the way Kelly describes it does sound like people who are not too skilled could still contribute. Is anyone else here able to do a little programming?”
There was nobody. Most people in the class were musicians.
“I am sure other people in the school will catch on and help”, Kelly said, “but it would be nice to do something in this class. I can’t do much myself, though.”
“Cleta and I don’t play music, Mr. Everett, so we don’t contribute much to this class. Could you ask the computer to swap us out and get a couple of programmers, instead?”, one student suggested.
The very next day the class reformed first thing in the morning with two good ST programmers replacing the two non-musicians in the group. Mr. Everett then had Kelly tell the class in more detail what they thought the class might do.
“The big program everyone uses is going to get broken down and redone, I’m sure, probably in ST, I hope. Even then it’s going to be too big for me to understand. We could help by splitting off a little piece to work on. But I also think a small program could be written to do some of the same things, then we could just make it bigger and better, keeping the software easy enough for a kid to understand. ST is good for that. I got started already. Well, I got started both ways, top down and bottom up. My parent’s told me might be good. It’s been fun, actually. We might just carry on the same way. Those of you who are musical, rather than programmers, could write the top level help stuff which has to be part of every ST program. You have to say what a program is going to do before you can run it. You musicians will know what it should do.”
A good idea, the class immediately adopted it. It remained a class project for exactly two weeks, until school principal Paul Grey got a demonstration.
“Good. You can be the lead class for now, and we can add on to what you have done, but I think this is now a school project!”
The class had mixed feelings about this. On the one had they felt their project had been hijacked. On the other, they were proud of getting it started. Kelly and Suzy felt just as ambivalent about it, until their parents praised them for their initiative. That did make them feel better.
The school offered dozens of small music classes and some larger ones. It offered even more computer classes. There were classes where people of extreme compatibility worked together. While still able to choose their own projects, most chose to participate in the bigger project.
The unique closely connected environment of Social Tech High made the success of this project inevitable. As the software got better and better, so did the music its Markov mechanisms generated from existing libraries of music.
The school still enjoyed connections to many other people from outside the school. Many of these were long-distance connections. Included were Beth Green, her artistic sister Esmeralda and Alice Ames. It was only natural that these people join in the school project. The fact that it was the school’s project never seemed to matter. These were school people, no matter how far away. If they could help, all the better.
The only stipulation was that all software had to be written in the school’s ST language. That language was designed to be easily taught to young students, easy to maintain, easy to document, yet powerful enough for all purposes.
Alice Ames had been part of the Social Tech project which had defined the language and written its first interpreter. She had written its first compiler, after rewriting an improved version of the ST interpreter in ST itself. Now an MIT student, Alice still used the language sometimes, though the university wanted her to code in other languages. Sufficiently fluent in ST and as brilliant as always, she loved her former school and would do anything for it.
Most fluent in the language were seniors who had been in the school since Grade Seven and done all their programming in it. Several of those skilled programmers were also musicians. They took over the project and led it well, but Kelly and Suzy still got to contribute.
Not everyone was happy with the rapidly developing software and the music generated by it. It soon seemed that everywhere you went there was music playing. That bothered a few people, who liked peace and quiet, but the really frustrated people in the school were those who liked to write songs or compose music themselves.
What is the point?”, Tennison Daley asked, sadly one day, while speaking to Kelly. “The machine turns out better music than I do.”
“I don’t know, Tenny. I just have this intuition, this feeling, that somehow what you do is important. Let’s ask someone.”
Kelly took the problem to Ann, who was like a favourite aunt. “Tennison writes music. Songs and stuff. Or at least she always did. Now she doesn’t see the point.”
“Ah. That problem. I have talked about it with Alice and some others. We got some good comments from Beth Green and her sister Esmeralda. It seems that Esmeralda is not just a visual artist, she does music too. I think I can explain it to you, maybe. Let me try. Does the notion of abstract space mean anything to you? What about hyperplane?”
“Yes, Ann, you see, in math class we spend more than half our time talking about math itself, less on specific things. We’ve been taught about abstract spaces of many dimensions. And hyperplanes are just like planes in our three dimensional world. They have one fewer dimension.”
“Well, close. There is the trivial hyperplane which is the whole abstract space itself, and a canonical hyperplane, which has just one fewer dimension, and the null hyperplane, with no dimensions at all, but a subspace with any number of dimensions can be a hyperplane.”
“OK. I guess.”
“So, think of Suzy’s playlist and the music generated from it. All classical sounding. Now suppose we deleted everything by Beethoven from her playlist. What would the generated music sound like?”
“Real different. Beethoven was big.”
“Exactly. Now think of Suzy’s playlist defining an abstract space of music. Without Beethoven, you get only a hyperplane, embedded within that abstract space. Removing Beethoven has removed a whole dimension.”
“I get it! I get it! So a human author contributes new dimensions to what would otherwise be confined to a hyperplane.”
“Right. So Tennison, what you should do is keep on writing you own original material. You can listen to what is already being produced and add something new, or just start from scratch. I’d say that you should do a lot of listening first. Listen to what the school produces and what your friends produce. Listen to what your own playlist of favourite pieces generates. Most especially listen to your the pair-music generated with your most compatible friends. Then try to create something new. Well, I recommend listening first, but what do I know? Do whatever works best for you. Kelly, we need a method by which people can add new pieces to their playlists. A way of submitting pieces so that people will actually listen to them and add them to their own collections.”
Kelly thought and thought and thought. Finally she brought an answer to her mother. Sally called in her friends Ellen Smith and Alma Renwick, then had Kelly repeat the idea.
“Well, we’ve stopped playing school music over the loudspeakers – that was kind of annoying. Now it is on a broadcast channel for in-school use. We can tune in and listen to it. Lots of us do. But for new music, I think we need another broadcast channel. Kinda like a school radio. Something to play only new music, written by real people. Not necessarily our students, but it could be. No, best we don’t limit it. Just real people anywhere. Then kids can download the good stuff into their collections.”
Kelly also repeated the conversation she’d had with Ann, summarizing her suggestion. It did seem like a good idea. Vice-principal Ellen Smith said she would talk it over with Paul Grey, but was sure she could make it happen.
It did not take long before there were a several wi-fi channels for different kinds of new music by real people. A school radio station accepted donated music, sometimes by students, often by artists from around the world seeking recognition in the hope of future fame and fortune. As Social Tech High spread the use of better and better software to generate music, it also became a clearing house for freely redistributable music by human artists.