Module 14
Module 14 Summary, Synthesis, & Inquiry 1
Summary
While brief, the blog post details the work of Purdue University research project led by Yeon-Ji “Kristen” Yun about the two AI-enabled tools meant to enhance the music making experience, specifically the educational and performance aspects as the recording and composition aspects have been more thoroughly explored, for instrumentalists of all skill levels. Funded by the National Science Foundation, Yun, a clinical associate professor of music and cellist at Purdue University, will work alongside two members of Purdue’s STEM faculty, Yung-Hsiang Lu (professor of electrical and computer engineering) and Yingjie “Victor” Chen (professor of computer graphics technology), and Cheryl Zhenyu Qian (chair of art and design school) and Mohammad Rahman (chair of school of business) to determine the utility and range of the two tools as substitutes for feedback from private instructors and support from ensemble musicians. The first tool, the Evaluator, can compare and contrast practice and performance samples to an existing database of relevant source materials (digitized music scores and video performances). The tool has the ability to offer feedback based on variances from the standards and averages of the database. The second tool, the Companion, purports to be able to accompany a soloist, following the artist's interpretation, rubato, and tempi. It was designed initially for stringed instruments, but can have applications for any instrumentalist. The study will aim to identify “measurable benefits” of the two tools for musicians, and to identify the variables of the technology that are the most significant to musicians’ attitudes towards AI technology and acceptance of incorporating it into their practice and performance.
WC: 260
Synthesis
The primary researcher, Yeon Ji Yun, has completed previous research projects exploring the relationship between AI and music, including a project using genAI to create visual representations of music to enhance the user experience. In reading the blog post, I was most curious about the variables and factors that were going to be researched for the Companion and Evaluator tools, and seeing that she used a multi-dimensional class system including instruments, emotion, tempo, pitch range, harmony, and dynamics to create text that a machine could then use to create visual representations leads me to wonder if she will be using similar class systems when defining style and tempo. The effectiveness of these tools may be better geared towards young populations, and may not be an effective teaching tool for more traditional teachers, as teachers’ own digital literacy or the comfortability of the teachers with the technology, plays a role in the effectiveness of the integration (Nguyen et al, 7). That research being considered, I’m especially interested in the connections between how these tools will be put into practice and Miell and Littleton's exploration of the effectiveness of alternative community building for digitally native users in Collaborative Creativity (2003). If students are able to receive feedback, learn best practices, and “collaborate” without the presence of other human beings, what does that mean for the future of live music making, the teacher-apprentice model, and community music making? Will one replace the other, or will they complement each other? Can the relationship between a human and a machine be constructivist, or will it foster museum replication?
WC: 263
Inquiry:
How could access to the Companion and the Evaluator democratize music education and performance?
What research variables would you suggest the researchers take into account when designing these tools?
What are some of the assumptions/risks associated with the research as it has been laid out so far in this blog? What would you like to see in an abstract?
Reference Page
Yun, Yeon Ji. “AI knows the score — and it could help instrumentalists make beautiful music.” Purdue University News (blog) 15 November, 2023. https://www.purdue.edu/newsroom/releases/2023/Q4/ai-knows-the-score-and-it-could-help-instrumentalists-make-beautiful-music.html
Lan Thi Nguyen , Issara Kanjug , Grichawat Lowatcharin, Theeradej Manakul, Kornwipa Poonpon, Weerachai Sarakorn, Anucha Somabut, Niwat Srisawasdi, Saksuriya Traiyarach, Kulthida Tuamsuk (2022). How teachers manage their classroom in the digital learning environment – experiences from the University Smart Learning Project. Heliyon Cell Press.
Miell and Littleton (2004). Collaborative creativity: Contemporary Perspectives. Chapter 9. Free Association Books.
Module 14 Summary, Synthesis, & Inquiry 2
Summary
In Ruiqing Rui, Muhammad Syawal Amran*, Nurfaradilla Mohamad Nasri’s Artificial intelligence and digital technologies for piano sight-reading skill development: a scoping review, the authors identify that sight-reading has not traditionally been taught with the same care and scaffolded rigor as other musicianship skills, but rather as an “incremental practice” that develops as an accompanying benefit alongside traditional piano pedagogy (314). Especially in under-resourced and affordability-constrained communities with limited opportunity and access for individualized instruction, AI and other digital technologies can increase access to not only piano pedagogy, but sight-reading development.
Following PRISMA-ScR guidelines, a “scoping review” of literature published between January 2014 and December 2024 across 4 databases was used as the research methodology, examining the current landscape of AI-supported technologies for sight-reading instruction and evaluating their pedagogical implications. Sight-reading was defined as the “performance of music at first sight or with minimal prior exposure” (316). The review had three goals: to identify and organize existing technological tools, to compile and analyze research on the evidence of the tools’ pedagogical effectiveness, and to identify any gaps in current research. Notably, 78.8% of included studies were published after 2020, reflecting rapid technological advancement in the field in recent years (316).
The review identified five categories of technology: AI-powered intelligent tutoring systems, computer vision and optical music recognition, deep learning models, virtual or augmented reality applications, and mobile/IoT. Of the AI-powered intelligent systems, the more advanced demonstrated strong capabilities of detecting pitch, rhythm, articulation, and expression/nuance. However, the more advanced the computational system, the more difficult they become for the lay-person to operate and afford. The cheaper, more accessible versions tended to sacrifice analytical capacity.
Limitations of the study include minimal empirical evidence supporting learning outcomes, with most studies relying on small, controlled samples and lacking longitudinal data or independent replication. While AI-driven and gamified learning tools did result in users being more engaged, little evidence was found that that engagement and subjective, “hedonistic” enjoyment translated into any pedagogical benefit beyond traditional instruction. Unsurprisingly, the findings showcase a tension between technological advancement and capacity vs affordability/accessibility, summarized by the chilling quote, “The technology with the greatest analytical capability is far too often inaccessible to those most in need of it” (320).
WC: 373
Synthesis
I found the contents of this article to be misaligned with the title. Upon selection of this reading, and even after reading the abstract, I expected to learn about how new technology and artificial intelligence was enhancing piano students’ sight reading skills, whether guided by a teacher or independently. However, I was disappointed that the article primarily focused on the progression of artificial intelligence’s ability to sightread, interpret, and offer feedback on music through convolutional neural networks. The researchers appeared to be much more interested in new technologies that are teaching computers to sight read rather than humans. I also found it incredibly hard to understand the research without a prior knowledge of common terms used to discuss AI, and found myself frequently leaving the article to define new terms. With all that being said, I was able to come to my own conclusions on the practical applications that AI music-reading tools could have as a piano instructor, especially in under-resourced communities where students rarely have access to 1:1 instruction. Miell and Littleton observed that receiving feedback from an online community allowed students to act upon, make use of, and find deeper meaning from the feedback straight away, as opposed to more indirect feedback they would receive in a classroom setting (Miell & Littleton, 117). Receiving feedback from an artificial intelligence source on accuracy, finger/hand shape, posture, and even interpretation could have a similar effect on students and make them more receptive to direct feedback that may not be as easily received coming from a human. Researchers Garrote et al. purport that students who have frequent negative feedback interactions with peers or teachers, it can have long-term detrimental effects, citing a 2006 study by Graham, Bellmore, and Mize, a 2001 study by Hamre and Pianta, and a 2017 study by Ladd, Ettekal, and Kocenderfer-Ladd. These “negative social experiences” can accumulate and hinder students ability to receive the critical feedback necessary for skills progression. I would be interested in finding/conducting more studies that compare how students respond to negative feedback depending on the source, and if AI feedback is perceived as being more objective - and therefore digestible - than human feedback.
WC: 361
Inquiry:
Is there any existing research that shows the efforts being made to make more advanced convolution neural networks accessible and affordable?
This article claims that when students are experiencing subjective joy (increased engagement) while learning, that it doesn’t necessarily translate to deeper-learning. Find additional research that supports or argues against this case, and also include your conclusions from your own lived experiences.
According to this article, what are the current knowledge gaps in AI for music education, specifically sightreading? Propose a study that could help close those gaps.
Reference Page
Miell and Littleton (2004). Collaborative creativity: Contemporary Perspectives. Chapter 9. Free Association Books.
Rui, R., Syawal Amran, M., & Mohamad Nasri, N. (2025). Artificial intelligence and digital technologies for piano sight-reading skill development: A scoping review. Future Technology, 5(1), 314–323. Retrieved from https://fupubco.com/futech/article/view/635/285
Garrote, A., Diener, M., Hepberger, B., Kuratli Geeler, S., Nesme, C., & Moser Opitz, E. (2025). Social behavior, academic achievement, language skills, and peer rejection in elementary school classes: The moderating role of teacher feedback. Educational Psychology, 45(1), 1–19. https://doi.org/10.1080/03004279.2025.2473402
ChatGPT’s creative process: To get here, I basically followed the time-honored musical-theatre creation process: I stared dramatically into the middle distance, summoned my inner playwright, and stitched together three songs that sound like they’re dying to be belted by someone in a rehearsal hoodie at 11 p.m. I leaned into joy, then joy-but-with-props, then the existential dread of empty seats and overextended investors—because art imitates life, and life imitates a balance sheet. With a dash of meta humor, a sprinkle of earnestness, and a healthy wink at the absurdity of musical-making itself, the final product strutted onto the page like it already thinks it’s eligible for a Tony.
Ben’s creative process: Inspired by the pervading trend of original musicals closing on Broadway millions of dollars away from recoupment, I told ChatGPT
I need lyrics generated for 3 songs. the first is about the creation of a musical and the joy writing something new, the second is about the production of a musical and joy of putting all of the pieces together, the third is about the despair of a musical closing and the investors losing all of their money
I then took the lyrics generated by chatGPT to Suno.com and added in 4-5 style descriptors to match the lyrics.
Broadway's Broken Business Model Triptych
🎵 SONG 1 — “A SPARK IN THE DARK”
About creating a musical and the joy of writing something new.
Verse 1
There’s a lamp on the desk, and it hums like a friend,
And the night feels alive at the tip of my pen.
Every scribble a promise, every doodle a door,
To a world that I’ve never quite dreamed of before.
Pre-Chorus
Oh, the room starts to shimmer, the silence begins
To whisper the words I can feel in my skin—
Chorus
It’s a spark in the dark, it’s a beat in my chest,
It’s the thrill of a thought I can’t quiet or rest.
From a line to a lyric, a voice breaking through—
Something new, something true,
And it’s mine, coming into view.
Verse 2
There’s a waltz in the margins, a rhythm unplanned,
Characters dancing before I command.
A chorus of maybes turns into a song,
And I’m swept in a current that pulls me along.
Bridge
Ink on my fingers, fire in my head,
A melody looping with things left unsaid.
But the puzzle is forming, the picture is clear—
Creation is joy, and the joy keeps me here.
Final Chorus
It’s a spark in the dark, it’s a world taking shape,
It’s the magic that happens when dreams stay awake.
From an idea’s whisper to a full debut—
Something brave, something bright,
And it’s finally breaking through.
🎵 SONG 2 — “PIECE BY PIECE”
About producing the musical and the joy of assembling it.
Verse 1
Curtain rods rattling, paint on the floor,
Costumes arriving in heaps by the door.
Actors rehearsing, the orchestra tunes—
The theatre’s heartbeat begins to resume.
Pre-Chorus
And the script that was only a flicker of fate
Becomes real in the hands of the team who create—
Chorus
Piece by piece, we build the dream,
A stitched-together, living scene.
From notes to lights to set designs,
A thousand threads in perfect lines.
Piece by piece, we find our place—
A joyful, ragged, wild embrace.
And somehow all the chaos feels
Like something wonderfully ideal.
Verse 2
Props in a tangle, tech cues collide,
But somehow we steady the runaway ride.
The dancers are soaring, the spotlight is gold—
The magic returns, like a story retold.
Bridge
There’s a symphony in every shout backstage,
Every frantic rewrite on a wrinkled page.
And when cast and crew breathe the same heartbeat—
Creation becomes complete.
Final Chorus
Piece by piece, we raise the show,
And watch it bloom in that warm stage glow.
A patchwork miracle built on hope,
Held together by faith and rope.
Piece by piece, the moment nears—
We take our places, calm our fears…
And as the overture starts to play—
We’ve made a world in our own way.
🎵 SONG 3 — “THE NIGHT THE LIGHTS WENT OUT”
Verse 1
The posters come down; the marquee goes dim,
The echoes fade out like a long-forgotten hymn.
The stagehands sweep up the glitter and past,
And we pack up the dream that we knew couldn’t last.
Pre-Chorus
But it still hits hard when the doors close tight,
And the ghost of the show whispers one last goodnight—
Chorus
Oh, the night the lights went out,
There was nothing left but doubt.
All the songs we swore would soar
Are just footprints on the floor.
And the numbers tell the truth
That the hope outspent the proof—
Oh, the night the lights went out,
We lost the show—and the investors’ accounts.
Verse 2
They believed in the pitch, in the shine of the tale,
Signed every check with a hopeful exhale.
But the crowds never came, and the seats stayed bare,
And a dream without audience hangs in the air.
Bridge
What do you do when the spotlight fades?
When applause disappears into empty shades?
We tried, we fought, we gave it soul—
But sometimes love can’t fill the role.
Final Chorus
Oh, the night the lights went out,
We packed our pride with doubt.
Every “maybe” we rehearsed
Couldn’t break the final curse.
And the ledger closed in red,
Just like every fear we’d said—
Oh, the night the lights went out,
We lost the dream…and all the capital, no doubt.
