 |
|
Может ли машина быть композитором?
| |
| ialexkalinin | Дата: Воскресенье, 17 Января 16, 03:51 | Сообщение # 1 |
|
Группа: Пользователи
Сообщений: 11
Статус: Offline
| Английские ученые решили открыть новый журнал Journal of Creative Music Systems (http://jcms.org.uk/), в связи с чем в некой узкой конференции возникло обсуждение. Тексты некоторых реплик оттуда привожу
-----------------------------------
From: Kent Walker
Sent: 11 January 2016 12:42
To: Valerio Velardo U1370329
Subject: Re: CfP - Inaugural Issue of the Journal of Creative Music Systems
For the life of me I don't understand how there can be a "scientific" journal dedicated to furthering Kurzweil's pseudo-religious beliefs in technology."Creative music systems" is truly religious propaganda.Most certainly there are NO computer systems capable of generating music. A "scientific" list that remains quiet on such issues is frankly unscientific.
-----------------------------------
Dear Kent,
I understand your concerns about creative music systems, but saying that "most certainly there are NO computer systems capable of generating music" is simply not true. If you are interested in the topic, the survey by Fernandez and Vico (2013) offers a quick review of hundreds of systems capable of generating music. As you might discover while reading the paper, algorithmic composition has a quite long history. The first creative music systems appeared in the late fifties. Today, automatic generation of music is a topic of interest in the most prestigious conferences on Artificial Intelligence (e.g., AAAI, ECAI). Algorithmic composition is but a subfield of Computational Creativity (CC) – the art and science of developing systems that are able to generate human-like creative artefacts such as painting, music and poetry. CC is a growing field with a large research community. Have a look at the Prosecco website if you want to know more (http://prosecco-network.eu/). I might be wrong, but the scepticism around computer creative systems resembles the scepticism around chess programmes 30 years ago. But then Deep Blue came…Fernández, J. D., & Vico, F. (2013). AI methods in algorithmic composition: A comprehensive survey. Journal of Artificial Intelligence Research, 513-582. https://goo.gl/xepRmf
Valerio
-----------------------------------
While I’m quite happy with the study of synthetic creativity, what I don’t understand is why you would want autonomous systems to write music. What is the purpose of it?
As a musician (of sorts) I can well understand systems to help me write or perform music, but why I would want a system to relieve me of my need to be creative I do not understand. If all I want is music to listen to, there’s lots of it available (and that’s a different problem, also causing pain to lots of performing musicians).
Or is it for autonomous synthetic listeners to listen to? That would be quite interesting: autonomous systems creating music, being listened to by an autonomous synthetic audience. You wouldn’t even need to actually play it, or even to convert it from digital to analog: it could be directly piped digitally from the synthetic composer to the synthetic audience, thus removing any requirement for human participation at all.
Yours (not too seriously)
Leslie Smith
----------------------------------------
Dear Kevin,
Thanks for the reply. I decided not to send this communication to the list. It was copied to the list via Valerio's reply (my fault). I agree that there a tone of intolerance displayed by my communication for which I have already apologized to Valerio. Yes, Nerve. Apologies to the list – and to the internet(s) for accidental exposure to my impolite exchange. Valerio and I shook digital hands and I think we're OK.For those not familiar with Kurzweil's pseudo-religious Ideas I suggest watching the film "The Singularity". http://www.imdb.com/title/tt2073120/They are self-admittedly religious. In the film there he says something like (paraphrase) "People ask me if there is a God, my reply to them is that there will be." He believes that what he refers to as "the singularity" is God, and that it will offer eternal life. Kurzweillian language (and there are others that preceded him ) has penetrated more deeply into the field of engineering than we admit. For example, even "machine learning" or "machine listening" are terms that have become quite common. To my knowledge we do not understand what learning is and although the brain operates using pattern recognition there many more factors than pattern recognition involved. Many of these factors are "multimodal" "embodied" and "ecological" (to use terms from Gibson – one of Dr. Bregman's influences). We should refer to these processes as "machine-simulated learning" and "machine-simulated listening".
I don't believe autonomous composition or algorithmic composition itself to be propaganda. Propaganda is the notion, rather, that the computers are doing the composing, that the computers themselves are creative, that we have realized artificial intelligence. In fact, there is not a clear agreement on what intelligence is. That is –as far as I know- the scientific description of reality. The reality is that composers are human, that music is a social practice. Composers eat, speak, make love, have being. Computers are not [yet?] capable of this. Sincerely,
-Kent
-----------------------------------------------------
Well, yes, I’m not actually going to take you very seriously at all. <<8-))>>> However I will try to articulate an answer to your question . . .
How can I take you seriously? I didn’t read the word ‘autonomous’ in the posting, or the word ‘write’. I did see the word compose.  Not being a computer scientist, but being a composer, I have been using [analog]computers for 45 years to invent / create / compose [sic], sounds. [Sorry about not using the word 'music’.]
My understanding of algorithmic composition is the model:
noise > filter > output
The interesting part of this model is the ‘filter’ — or more practically, filters. The noise source generates a [quasi]random sequence of numbers [voltages / energy levels]. The filters select / shape the acceptability of a number to fit some kind of information mask. The information mask in the analog synthesis domain is loosely called a ‘patch’, or in the hybrid / digital worlds, an algorithm.
I have done this in various ways for decades. The purpose? For me, to come to a better understanding of how my mind works. In my life, I view understanding of how the brain receives chaotic information and filters it into ‘meaningful’ form to be quite valuable. In one sense, it is a kind of ‘education’. And I’m a big fan of education. Education is like lighting a match in a dark cave.
There are those who simply ‘want the experience’, and they can ask Siri the question. And Siri will give them the answer. Or perhaps an answer. Or perhaps not.
My generalized teaching method — the match again, is to start with an idea, build models of how the idea could be understood — called analysis, derive possible ‘rules’ from / for the analysis, and to find out if the model [analysis]and rules are accurate and [adequately]complete. These rules are used to synthesize a response.
This could be a class of five year olds learning how to improvise on Orff instruments, or it could be music of the new complexity. The fundamentals, for me, remain the same.
For me, that is the purpose. <<8-()>>
And also to fill up the time between my first and last breath.
Kevin
|
| |
| |
| nibelung | Дата: Воскресенье, 17 Января 16, 12:50 | Сообщение # 2 |
Группа: Проверенные
Сообщений: 2281
Статус: Offline
| Может. Я же композитор!
|
| |
| |
| theodor19 | Дата: Воскресенье, 17 Января 16, 14:00 | Сообщение # 3 |
Группа: Модераторы
Сообщений: 1089
Статус: Online
| Еще как может.
Ещё и стихи сочинять может - см. С.Лем, "Кибериада".
Вот:
Лопотуй голомозый, да бундет грывчато
В кочь турмельной бычахе, что коздрой уснит,
Окошел бы назакрочь, высвиря глазята,
А порсаки корсливые вычат намрыд!
Скабичевский, – пропищал тот, почему-то указывая на свой примус.
|
| |
| |
| nepman | Дата: Воскресенье, 17 Января 16, 14:54 | Сообщение # 4 |
|
Группа: Пользователи
Сообщений: 170
Статус: Offline
| Зачем еще и компьютеру чего-то сочинять, когда и так сочиняют тыщу сочинений в год всякими композерами...
|
| |
| |
| ialexkalinin | Дата: Воскресенье, 17 Января 16, 15:06 | Сообщение # 5 |
|
Группа: Пользователи
Сообщений: 11
Статус: Offline
| Цитата nibelung (  ) Может. Я же композитор!
//и стихи сочинять может
Может быть. Но тут о другом. Народ интересуется - зачем де музыку будет писать машина, а не человек? Не дает ответа Валерио. Вот он этот начинающий доктор Фауст:
https://www.youtube.com/watch?v=ZfdoddYhkYA
|
| |
| |
| bulaguero | Дата: Вторник, 19 Января 16, 15:50 | Сообщение # 6 |
Группа: Пользователи
Сообщений: 56
Статус: Offline
| Вспоминается фраза профессора консерватории из фильма "Приходите завтра": "Можно, наверное, и осла научить петь. Но зачем?! А впрочем, мне это никогда не удавалось..."
|
| |
| |
| gellao5 | Дата: Вторник, 19 Января 16, 19:14 | Сообщение # 7 |
|
Группа: Проверенные
Сообщений: 221
Статус: Offline
| Конечно может. Но каким...
|
| |
| |
| Земля | Дата: Среда, 20 Января 16, 09:12 | Сообщение # 8 |
Группа: Проверенные
Сообщений: 2962
Статус: Offline
| Сможет. Если захочет. Добавлено (20 Января 16, 09:12)
---------------------------------------------
А вот насильно заставлять не надо бы, товарищи. Это просто негуманно!
|
| |
| |
| Cassis | Дата: Суббота, 23 Января 16, 15:37 | Сообщение # 9 |
Группа: Проверенные
Сообщений: 2329
Статус: Offline
| может ли женщина стать композитором? на этот вопрос так и не найден ответ, а вы сразу за машину берётесь [набросил маленько]
Mon centre cède, ma droite recule, situation excellente, j’attaque. Ferdinand Foch
|
| |
| |
| sidmak | Дата: Понедельник, 25 Января 16, 01:01 | Сообщение # 10 |
Группа: Модераторы
Сообщений: 895
Статус: Offline
| Может ли машина?
Да, да, да, да, да!
И ещё замечу, что могла всегда!
Ритмы шестерёнок...
Тихий шелест шин...
Просто плохо знаем музыку машин! :)
https://www.youtube.com/watch?v=deGZI7mc0BE
|
| |
| |
|
|
