Make AI sing an Enka song? ACE Huolian(AI singer)★Yoko Nagayama «Koi-no Tsugaru Jūsanko/恋の津軽十三湖»

Details
Title | Make AI sing an Enka song? ACE Huolian(AI singer)★Yoko Nagayama «Koi-no Tsugaru Jūsanko/恋の津軽十三湖» |
Author | Wendi Xiao |
Duration | 2:14 |
File Format | MP3 / MP4 |
Original URL | https://youtube.com/watch?v=yPyd2BBMo8g |
Description
This cover of the traditional Enka Song «Koi-no Tsugaru Jūsanko / Lover on Tsugaru's lakeside» is completely sung (produced) by AI rather than by a real human singer. As the operator, I inputted simple musical instructions (linear melody, syllables of lyrics, articulations etc.) to the AI system of ACE Xuni-Geji 2.0 (ACE 虚拟歌姬2.0), and chose one of the best versions synthesized, automatically and internally, which met my (an ordinary human‘s) aesthetic taste as a final output.
AI enables us to project our thoughts and desires into a non-real wonderland. While music, as the catalyst, dissolving the boundary between the virtual worlds and the real ones, makes us interact with our avatars or digital doubles. In the dreamland where people and machines cohabit evenly, new forms of emotional connection and humanistic care are gradually blooming.
REAL vs VIRTUAL is an old topic as cliché as «imatation». The music generated by AI is undoubtedly a typical product from this theory source. However, rather than some idealized obsession with transcendence or fanaticism for formalism, I tried to answer a new question with the help of the artistic creation by the virtual body : can post-humain develop some new aesthetic principles based on the virtualized presentations, in order to solve some inner, innate, and irreconcilable contradictions of our contemporary artistic life?
The answer becomes clear as more neophytes participate in finishing the blueprint of metaverse. Here, in order to discuss the autonomy of AI in creating art, I prefer to proceed from the auditory experience of this AI Enka Song to verify a presupposition as follow:
«In the future, the new-art will be completely constructed in the form of a mixture of virtual and real, and man-machine collaboration will be the source of methodology and bring theoretical norms that transcend the limitations of the physical world.»
As proof, this virtual Enka symbolizes a new life form - a «perfect subject» of imitated truth. That is to say, AI makes us enable to hear such an extreme case in which an Enka could be sung extraordinarily Enka-ly. Exagerated vibratos, microtonal voicings, sobbing tune, wheezing… everything appeared more natural than those of a real human diva. I can even hear the provocation of the AI:
«we have already crossed the threshold of your human fragile physical body.»
So, well, in this piece, «immortality» is being realized by AI. With the improvement of algorithm and computing power, convolution neural network technology, which imitates human thinking habits, has become a normal logic of artistic creation. Music abandons the human body, transcends the binary defines like time-space or mind-body, and becomes the digital «beauté libre». I didn't intend to conceive the «homo-deus» in a musical way, but just locate myself in a métarécit of post-human, being eager for granting that the immortal virtual body have the same legitimacy as the human body. Furthermore, they will be more valorous to wandering freely in the metaverse.
Meanwhlie, I seemed to have succeeded in updating a philosophical paradox (or just a jeu de mots):
VIRTUAL songs are more real than any REAL songs.
A salvation? An SOS? Or an oracle? Only we know is that the contemporary human body has lost its sensitivity under the stimulation of electronic waves and digital eddies. The situation of continuous repression and longtime aphasia makes us unwittingly transfer our subjectivity to the inhuman «autrui». Our understanding of existentialism is no longer valid, thus we are more confused about the infinite freedom. When the so-called avant-garde spirit no longer has a positive moral significance, where should our contemporary music go?
Maybe the virtual bodies know more than we do.
*** *** *** *** ***
I would like to thank the original singer of this song, Mme. Yoko NAGAYAMA, one of the most famous Enka divas in Japan, and the Chinese technology company ACE Studio, which provides me with the technical support and its open source experimental AI operating platform.