by Ivan Linn
Speaking of AI, even individuals who are indifferent to the field of science and technology can perceive the profound impact it has exerted on our society and its immense potential. This “enigmatic” force is reshaping various facets of our world, including the music industry. Prior to delving into this subject matter, it is imperative for us to grasp the fundamental concepts and historical background pertaining to both music and AI.
1. What is Music? What is AI?
When discussing music, what is the subject matter? From a scientific point of view, music is generally defined as the art of arranging sound to create some combination of form, harmony, melody, rhythm, or otherwise expressive content.[1] In layman’s terms, music is an auditory art form that encompasses aesthetic beauty, introspection, and a profound appreciation for life. Music represents the essence of humanity. The historical origins of music can be traced back to ancient civilizations where primitive instruments like hunting horns and drums were utilized as means of communication or emotional expression.
The music of prehistoric cultures is first firmly dated to c. 40,000 BP of the Upper Paleolithic by evidence of bone flutes.[2] It can be argued that music emerged concomitantly with the advent of human civilization, making it as ancient as the genetic makeup of humanity itself. In contemporary society, music has permeated every individual’s senses, evoking profound emotions and becoming an integral part of modern civilization.
As for AI, the epitome of this era’s scientific and technological advancements, what constitutes its most fundamental concept? Artificial intelligence (AI) is the intelligence of machines or software, as opposed to the intelligence of humans or animals.[3] The history of AI is relatively short.，in the first half of the 20th century, science fiction familiarized the world with the concept of artificially intelligent robots. It began with the “heartless” Tin man from the Wizard of Oz and continued with the humanoid robot that impersonated Maria in Metropolis. [4] The integration of AI into the realm of rigorous scientific inquiry has been a gradual process. 1950, Alan Turing introduced the Turing test in his paper “Computing Machinery and Intelligence”[5]. Summer of 1956, Dartmouth Conference launched the field of Al and coined the term “artificial intelligence”. And the first artificial intelligence program and was presented at the Dartmouth Summer Research Project on Artificial Intelligence (DSRPAI) hosted by John McCarthy and Marvin Minsky in that year.[6] In the 1980s, John Hopfield and David Rumelhart popularized “deep learning” techniques which allowed computers to learn using experience. On may 11, 1997, IBM’s Deep Blue computer beat reigning world chess champion Garry Kasparov. This highly publicized match was the first time a reigning world chess champion lost to a computer and served as a huge step towards an artificially intelligent decision making program.[7] On May 23, 2017, Alpha Go defeated Ke Jie, the world’s number one Go player at that time. This means that artificial intelligence has surpassed human intelligence in the strongest intellectual project of mankind.[8] On November 30, 2022, OpenAI, headquartered in San Francisco, announced the launch of ChatGPT, which triggered a global AI research and development and GPT model research and development craze.[9]
2. Encounter — the intertwining of destiny
Music and AI, as the two paramount art forms in human society, converge at the crossroads of historical progression under the orchestration of fate. In 1956, Lejaren Hiller, a composer born chemist, suggested that a computer, properly programmed, it could be used as a support tool for composing music.[10] The first artificial intelligence music dates back to 1958, which was created by Iannis Xenakis and is generally regarded as the first algorithmic piece ever written.[11] Xenakis used a very old technology which called Markov Chains to create this. Although the music created by this technology may be beyond the aesthetic range of ordinary people, it still cannot deny the historic significance of this attempt. In 1960, Russian researcher Rudolf Zaripov published worldwide first paper on algorithmic music composing using the “Ural-1” computer.[12] In 1989, more than thirty years later, the first neural network was used to create music by Markov Chains, and its breakthrough was to give music a real structure. In 2002, Doug Eck used LSTM neural networks to create music as a pioneer. And now, he has become head of the Magenta project at Google Brain while is considered one of the leaders in the music artificial intelligence field.[13] In 2012, one of the first examples of an AI-generated music program, Iamus, developed by researchers at the University of Malaga in Spain and produced its first full-length album which was titled “Iamus”.[14] In 2017, Popgun which Stephen Phillips is one of the founders have built an AI called Alice.[15] Alice is totally unsupervised and could understand what the human pianist is playing then plays along with him. This means its music AI technology was so close to today.
In recent years, AI music creation has made significant advancements due to breakthroughs in technologies such as deep learning and neural networks. For example, apps like Wavv MusicGPT, Stability AI, IBM Watson Beat, Amper Music and Soundtrap can create customized background music for videos or podcasts now. And AI-powered virtual musicians like “Magenta” and virtual bands like “The Annoying Orange” can create music performances and collaborate with human musicians.
3. People’s historical dynamic cognition of AI music
When discussing AI music today, there is an overwhelming sense of boundless anticipation and reverie. However, the amalgamation of AI and music has not always been met with such positivity.
In 1958, during the emergence of AI music, many individuals expressed novelty and astonishment towards it. However, due to its deviation from conventional musical preferences, it failed to garner recognition or even attention. Over the subsequent decades, despite continuous advancements in AI music technology, the quality of compositions produced remained significantly inferior to those crafted by human musicians and consequently lacked widespread acclaim. Particularly during the distinctive era known as the “winter of artificial intelligence” in the 1970s and 1980s, public interest in AI music dwindled considerably.
It was not until the 1990s that substantial progress was made in terms of AI’s ability to create music following breakthroughs in computer hardware and advancements in AI technology. During this period, certain aspects of AI-generated compositions began rivaling human musicians’ performances which captivated numerous researchers within this field. Nevertheless, hindered by factors such as high costs and limited application scenarios, AI music failed to make significant waves within mainstream society.
In recent years (2010s), with further developments in AI technology unfolding before us, people have gradually come to realize its potential within the realm of music. The number of researchers specializing in AI-generated compositions has increased exponentially while various commercial projects have successfully materialized — progressively capturing public attention along the way. However, alongside expressions of intrigue and acknowledgment towards these advancements emerged voices ridiculing or opposing them as well. For instance, some argue that a lack of genuine artistry or uniqueness renders AI-created pieces. There is also the contention that music is fundamentally an expression of emotion, and AI, being a machine, is incapable of comprehending or articulating emotions. Conversely, in recent years, the exponential advancement of AI technology has rendered the assertion “AI cannot do this” uncommon. Indeed, AI has surpassed human capabilities in nearly every domain, including the artistic expression of sentiment in music.
4. The prospects for AI music
Regarding the future development and combination of AI and music, we are both optimistic and curious. We can anticipate an increasing prevalence of AI applications in the domain of music as technology continues to advance. For instance, AI can assist musicians in gaining a deeper understanding of the preferences and requirements of the audience, enabling them to compose works of music that are more in tune with market demands. Moreover, AI can assist musicians in generating and executing real-time adjustments and feedback to enhance their performance.
In conclusion, the intersection of artificial intelligence and music presents both obstacles and prospects. Despite the fact that the public’s perception of AI music has experienced a series of dynamic shifts, we remain optimistic that AI will provide new opportunities and space for growth in the realm of music creation through ongoing investigation and advancement. Together, we should anticipate the future of AI and music.
[1] Oxford English Dictionary Online. https://www.oed.com/dictionary/music_n
[2] History of music-Wikipedia. https://en.wikipedia.org/wiki/History_of_music
[3] Artificial intelligence-Wikipedia. https://en.wikipedia.org/wiki/Artificial_intelligence#In_fiction
[4] The History of Artificial Intelligence，Rockwell Anyoha，Harvard University. https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/
[5] Brief Timeline of AI. https://www.livescience.com/47544-history-of-a-i-artificial-intelligence-infographic.html
[6] The History of Artificial Intelligence，Rockwell Anyoha，Harvard University. https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/
[7] Brief Timeline of AI. https://www.livescience.com/47544-history-of-a-i-artificial-intelligence-infographic.html
[8] Google’s AlphaGo AI defeats world Go number one Ke Jie.
https://www.theverge.com/2017/5/23/15679110/go-alphago-ke-jie-match-google-deepmind-ai-2017
[9] OpenAI finally introduces a business version of ChatGPT.
https://www.zdnet.com/article/openai-finally-introduces-a-business-version-of-chatgpt/
[10] Lejaren Hiller, Leonard Isaacson Experimental Music: Composition With an Electronic Computer.
[11] Artificial Intelligence, Making Music Since 1958: The Journey From Then to Today.
[12] Zaripov, Rudolf (1960). Proceedings of the USSR Academy of Sciences. 132 (6).
[13] Artificial Intelligence, Making Music Since 1958: The Journey From Then to Today.
[14] AI-generated music: a new soundscape. https://aicontentfy.com/en/blog/ai-generated-music-new-soundscape#:~:text=The%20history%20of%20AI%2Dgenerated%20music%20can%20be%20traced%20back,using%20computers%20to%20create%20music.
[15] Artificial Intelligence, Making Music Since 1958: The Journey From Then to Today.
