This article is part of my ongoing series exploring different dimensions of working with sound across creative and cultural contexts.
“ХИЙМЭЛ ОЮУН” (khii-mel o-yun) — or “XO” for short — is the Mongolian term for Artificial Intelligence, used in both everyday and academic contexts. On CPinMongolia.com this XO logo identifies articles about research and creative practices that explore how musician-creators are engaging with AI not as a replacement, but as a co-creator in their artistic expression.
______________________________________________
Without leaning into the broader debate around AI’s role in the future, I would like to share the following grounded example of how AI can be used by musicians attuned to aesthetics, meaning and guided by their own creative vision.
During my recent stay in Mongolia, I was struck by the way in which a new generation of Mongolian creative artists are embracing a diverse range of AI tools as co-creative partners in their artistic expression, blending tradition with technology in innovative and culturally resonant ways.
The following music video is a very good Mongolian-in-Mogolia example to open a discussion. The emphasis in the article is on the musical rather than the visual — put aside, for another time. For those of you not fluent or familiar with the Mongolian language (Cyrillic), the lyrics of the song have been translated into English* (see below). If there are any mistakes in this regard, they are of my own making, and for these I sincerely apologise to my Mongolian and other readers.
______________________________________________
Чандмань Эрдэнэ (Chandmani Erdene)
This is a very well-known traditional Mongolian song initially popularised by performers such as Adarsuren (also see below). As far as I can work out, there are no verified publicly available registers that clearly identify the lyricist (үг зохиогч) of this song in either English or Mongolian sources. If any of you know, please send me an email…
I first heard Chandmani Erdene in Mongolia many years ago. It moved me so deeply! Since then, its beauty has never faded — sometimes it brings tears to my eyes; at other times, it sends a quiet shiver up my spine. The effect is always visceral, touching something beyond language — felt rather than thought …
Song Lyrics (Eng. trans.)
The moonlight that shines in my eyes
The radiance absorbed deep in my trembling heart
Your beauty stirs flowing warmth within me
Your beauty outshines the gleam of crystal, even with a star at its side
Your grace softens the vast Khangaï and Gobi
With a warmth that could soften even stone
In this vast world, you are my warming Sun
Your radiant beams purify my open heart
Your beauty lights up a petalled mirror
With the reins to draw even the farthest near
Your joyful spirit delights the living Earth
And stirs all life toward its radiant flourish
Bound by reins, we move as one through fields in bloom
We moved with the melody in the joyful breeze— was it not us?
There in the murmur of the Tsenkher River’s flow
Echoed a song — was it not you who shimmered in its ripples?
Upon the snowy peak of Tsagaan Sairdag, my melody rests
Even in harsh winds, it was you who stood firm
The strings of my horse-headed lute pull taut
As the tsuur flute coaxes the melodies to align
There is a song that brought us together
Let its wholehearted melody keep dancing with joy
You, whose nature shines like a precious celestial jewel
Grant me your celestial blessing
You, whose nature shines like a precious celestial jewel
Grant me your celestial blessing
_________________________
Full disclosure: I love this latest version by Ai Cassette, especially for its thoughtful and culturally sensitive approach. With that in mind, here are a few general reflections from my reading of this elegant and moving artistic work.
Эмэртэй (emertei)
More than just AI output, this “emertei” cover reflects a young Mongolian creator’s unique taste, style, and expressive direction. This AI-co-created cover by a young Mongolian artist is truly “emertei” — infused with personal style, aesthetic flair, and a bold creative voice that goes beyond the algorithm.In the socio-cultural context of a music video created with AI by this Mongolian artist, the term “emertei cover” takes on a nuanced, expressive meaning beyond its literal translation.
“Emertei” describes a cover version that is stylised, emotionally charged, and visually expressive, often with a strong personal or artistic identity. In youth and creative circles, it suggests a work that carries flair, attitude, or signature energy — something that feels intentional, tuned to aesthetics, and not generic.
In this emerging creative space, where AI tools are accessible to (young) Mongolians experimenting with visuals and sound, calling a piece “эмэртэй” may signal that: (i) it’s not just generated, but curated or directed; (ii) there’s personality and taste embedded in the output; and that (iii) the artist has made conscious aesthetic decisions, even when using AI. The composition stands apart from flat or neutral outputs — it has style. So a phrase like “AI-гаар хийсэн эмэртэй cover” would be understood in cultural context as: “An AI-co-created cover with strong artistic flair or personal style.”
_________________________
Instrumental only version with on-screen lyrics
In terms of contemporary musical sociality, my experience and observations in Mongolia since 2004 highlight social singing as a vibrant, shared practice across diverse settings and age groups—from concerts and karaoke to more intimate personal and community gatherings. In support of this ongoing tradition, Ai Cassette has generously provided an instrumental-only with on-screen lyrics version for solo vocalists or groups who wish to perform the song themselves. Easy to do (but not necessary) for those who have done so much analysis and production work — hence my use of the term, ‘generous’. Thank you!
_________________________
Mongolian singer Adarsuren (Адарсүрэн)
Ai Casette’s 2025 music video (this version) connects to one of the most notable (and widely-loved) interpretations of “Чандмань Эрдэнэ” in Mongolia — that sung by Adarsuren, whose rendition offers us a distinctly heartfelt and measured male vocal tone that expresses and highlights the song’s lyrical warmth and poetic imagery.
The following video also provides a valuable example of how traditional Mongolian music ‘artefacts’ continue to be preserved and reinterpreted through digital platforms, reaching wider audiences (of their time) while maintaining originating culture’s authenticity.
__________________________________________________________________
NOTES
Notes on English translation choices
- The original lyrics move fluidly between present and past tenses, which is typical in Mongolian lyric poetry, where memory, presence, and timeless admiration often coexist.
-
For this translation intended for non-native speakers of Mongolian, I have omitted “Чи минь хө” and “хө”, the affectionate poetic markers often left untranslated and rendered as “oh you” for tone and rhythm in the original.
-
“Хангай говь” in the original is symbolic of Mongolia’s rich natural contrast — mountain and desert.
-
“Чандмань” refers to a mythical or celestial gem in Mongolian and Buddhist tradition — implying rarity, beauty, and divine power.
_________________
More Notes: 10 must-do tasks in AI co-creation
Motivation is the heartbeat of creative AI use—it’s what transforms mere algorithms into tools for designing and shaping something beautiful, whatever the artistic modality and its aesthetics. Bringing artistic vision to life with AI involves a surprising amount of discretionary work: countless subtle choices, refinements, and human judgment that shape the chosen aesthetic, the specific style, mood, or artistic direction that a person intentionally selects to guide the look, sound, or feel of their creative work.
In the context of music or AI-assisted art, it’s about making conscious decisions—like tone, genre, colour, rhythm, or emotional atmosphere—that reflect personal taste or artistic goals. Think of it as the deliberate “feel” you aim for, shaped by values, culture, and creative intent. And so, here are 10 must-do tasks, focussing on the musical and technical, that bridge AI tools and human creative intent:
1. Refine AI-Generated Audio with Manual Mixing
While AI tools can generate a decent mix, the real artistry lies in the human touch. Manual mixing—adjusting EQ, compression, stereo panning, reverb depth, and volume automation—is essential to bring a track to life. These decisions are deeply personal and context-sensitive, often requiring trained ears to sculpt the sound to match the emotional tone and style of the piece. Without this step, AI-generated audio can sound flat or unfinished, lacking the sonic balance and clarity expected in professional releases (Elkington, 2024; Johnson, 2024).
2. Inject Emotion Through Dynamic Arrangement
AI tends to produce static arrangements that don’t evolve emotionally in the way human-made compositions do. To truly express an artistic vision, it’s vital to shape the arrangement with deliberate dynamic shifts—building tension, releasing energy, and guiding the listener through emotional highs and lows. These changes require a strong intuitive sense of storytelling through sound (Ma, 2024; Elkington, 2024).
3. Replace or Layer AI Vocals with Real Takes
AI-generated vocals, while increasingly impressive, often still fall short in conveying deep emotional nuances or personalised expression. Replacing key sections with human-recorded vocals—or layering real takes with AI voices—adds depth, authenticity, and character. The subtle breathiness, timing imperfections, and natural inflections in a human voice can make a significant difference in how a track connects with listeners emotionally (Johnson, 2024; Elkington, 2024).
4. Edit MIDI & Velocity for Human Feel
AI-composed MIDI patterns can often feel too rigid or mathematically perfect, lacking the organic timing and expressive touch of a human performer. To address this, it’s important to go in and “humanise”—a common term and technical function used in digital musical production tools—the MIDI by slightly altering note timings and adjusting velocities. This brings a sense of groove and realism, particularly important in genres where rhythm and phrasing are central to emotional impact—like jazz or cinematic music (Ma, 2024; Elkington, 2024).
5. Curate & Customize AI Sound Design
AI-generated sounds—whether synths, textures, or instruments—often serve as a good starting point, but rarely match the uniqueness an artist might want. Customising these sounds by tweaking synth parameters, layering samples, or processing them through analog-style effects can help develop a sonic identity. These subtle sound design decisions are where much of the individuality and originality of a track emerge, distinguishing it from the generic tone of algorithmic defaults (Johnson, 2024; Ma, 2024).
6. Maintain Thematic & Tonal Consistency
One common pitfall in AI-assisted composition is inconsistency. AI may shift styles, keys, or textures without understanding the broader thematic intent. Maintaining cohesion across the piece—whether through repeated motifs, consistent harmonic language, or a unified instrumentation palette—is a human judgment call. This ensures the piece resonates as a coherent artistic statement, rather than a collage of disconnected ideas (Elkington, 2024; Johnson, 2024).
7.Perform A/B Listening on Multiple Systems
AI mixing tools often optimise for studio conditions, but real-world playback environments vary greatly. It’s essential to test the mix across multiple systems: headphones, car speakers, smartphone speakers, and high-end monitors. A/B listening helps catch frequency imbalances or phase issues that AI tools may overlook. This discretionary effort ensures the music sounds good wherever it’s played and is especially critical before publishing or distributing widely (Ma, 2024; Elkington, 2024).
8.Organise Iterations and Save Creative Decisions
Working with AI often involves generating multiple versions of a track or section. Without a disciplined system for organising these iterations, it’s easy to lose track of what worked and why. Keeping labelled project files, taking notes on creative decisions, and archiving versions that show progress allows you to refine ideas over time. This process not only preserves your creative momentum but also helps you learn what aligns best with your evolving artistic identity (Johnson, 2024; Ma, 2024).
_________________________
Further Reading
Elkington, R. (2024). Critical Making: AI in Screen and Audio Education. Queensland University of Technology. https://eprints.qut.edu.au/252460
This report explores the pedagogical and creative implications of using AI as a co-creative partner in audio and screen production education. It provides specific case reflections on student work, highlighting how human discretion and editing shape AI outputs in music-based creative practice.
Johnson, L. (2024). Xeno AI: How Can Creative Arts Practice Express Disalienation of an AI System? RMIT University. https://research-repository.rmit.edu.au/articles/thesis/Xeno_AI_How_can_Creative_Arts_Practice_Express_Disalienation_of_an_AI_system_/28657115
A practice-based PhD project investigating AI as a musical and expressive collaborator in sound-based creative arts. The work focuses on affect, identity, and agency within human-AI creative relations, making it directly relevant to co-composition and decision-making in sound.
Ma, S. (2024). Touching Wires: Exploring Tactility in Human-AI Musical Co-Creation. Australian National University. https://openresearch-repository.anu.edu.au/bitstreams/8cd348a5-0c18-4c79-82e1-a36b62a2ae01/download
This thesis examines the embodied and tactile aspects of co-creating music with AI, proposing that physical engagement and intuition remain essential even when working with advanced generative systems. It bridges creative practice with deep inquiry into musical choice, expression, and AI interactivity.
_________________________
Please refer to the INDEX for other music and articles that may be of interest.
End of transcript.
© 2013-2026. CP in Mongolia. “Letter from Mongolia 14: Chandmani Erdene” is licensed under the Creative Commons Attribution 4.0 International License. ScholarGPT provided an additional channel for research. Documents linked from this page may be subject to other restrictions. Posted: 11 February 2026. Last updated: 11 February 2026.