As one of many first college students in MIT’s new Music Know-how and Computation Graduate Program, Mariano Salcedo ’25 is researching the intersection between synthetic intelligence and music visuals.
Particularly, his graduate analysis focuses on neural mobile automata (NCA), which merges classical mobile automata with machine studying methods to develop photographs that may regenerate.
When paired with a stimulus like music, these photographs can “present” sounds in motion.
“This strategy allows anybody to create music-driven visuals whereas leveraging the expressive and typically unpredictable dynamics of self-organized programs,” Salcedo says. Via the online interface Salcedo has designed, customers can modify the connection between the music’s power and the NCA system to create distinctive visible performances utilizing any music audio stream.
“I would like the visuals to enrich and elevate the listening expertise,” he says.
Final 12 months Salcedo, the Alex Rigopulos (1992) Fellow in Music Know-how and Computation, earned a BS in synthetic intelligence and determination making from MIT, the place he explored sign processing in machine studying and the way a classical understanding of alerts can inform how we perceive AI. Now he’s one in all 5 grasp’s college students within the Music Know-how and Computation Graduate Program’s inaugural cohort.
This system, directed by professor of the apply in music know-how Eran Egozy ’93, MNG ’95, is a collaboration between MIT Music and Theater Arts within the College of Humanities, Arts, and Social Sciences, and the College of Engineering. It invitations practitioners to check, uncover, and develop new computational approaches to music. It additionally features a speaker collection that exposes college students and the broader MIT group to music business professionals, artists, technologists, and different researchers.
Rigopulos ’92, SM ’94, is a online game designer, musician, and former CEO of Harmonix Music Techniques, an organization he co-founded with Egozy in 1995. Harmonix is now part of Epic Video games, the place Rigopulos is the director of recreation improvement for music.
“MIT is the place I used to be first capable of pursue my ardour for music know-how a long time in the past, and that have was the springboard for a protracted and fulfilling profession,” says Rigopulos. “So, when MIT launched a sophisticated diploma program in music know-how, I used to be thrilled to fund a fellowship to assist propel this thrilling new program.”
Egozy is captivated with Salcedo’s work and his dedication to additional exploring its prospects. “He’s a gorgeous instance of a multidisciplinary researcher who thinks deeply about the way to greatest use know-how to reinforce and broaden human creativity,” he says.
Salcedo has been chosen to ship the coed deal with on the 2026 Superior Diploma Ceremony for the College of Humanities, Arts, and Social Sciences. “It’s an honor and it’s daunting,” he says. “It appears like an enormous accountability,” although one he’s wanting to embrace. His choice additionally pleases Egozy. “I’m tremendous excited that Maraino was chosen to ship this 12 months’s keynote,” he enthuses.
Altering gears
Rising up in Mexico and Texas, Mariano Salcedo couldn’t readily indulge his ardour for creating music. “There are not any bands in Mexican public faculties,” he says. Whereas some households may pay for devices and classes, others like Salcedo’s had been much less lucky.
“I’ve at all times liked music,” he continues. “I used to be a listener.”
Salcedo started his MIT journey as a mechanical engineering pupil, making use of to MIT by way of the Questbridge program. “I heard if you happen to like engineering and science that attending MIT could be a terrific selection,” he remembers. “Nerds are welcomed and embraced.” Whereas he dutifully labored towards finishing his MechE curriculum, music and know-how got here calling after an opportunity encounter with an LLM.
“I used to be launched to an LLM chatbot and was blown away,” he remembers. “This was one thing that was chatting with me. I used to be each awed and frightened.” After his encounter with the chatbot, Salcedo switched his main from mechanical engineering to synthetic intelligence and determination making.
“I mainly began over after being two thirds of the best way by way of the MechE curriculum,” he says. He realized concerning the prospects obtainable with AI but additionally confronted among the challenges bedeviling researchers and builders together with its potential energy, making certain its accountable use, human bias, restricted entry for folks from underrepresented teams, and an absence of variety amongst builders. He determined he would possibly have the ability to change that image.
“I assumed yet one more individual within the discipline may make a distinction,” he says.
Whereas finishing his undergraduate research, Salcedo’s love of music resurfaced. “I started DJ’ing at MIT and was hooked,” he says. Whereas he hadn’t realized to play a conventional instrument, he found he may create participating soundscapes with know-how. “I purchased a digital audio work station to assist me make music,” he continues.
Egozy and Salcedo met in 2024 whereas Salcedo accomplished an Undergraduate Analysis Alternatives Program rotation as a recreation developer in Egozy’s lab. “He was extremely curious and has grown tremendously over a really quick time interval,” Egozy says. Egozy grew to become an off-the-cuff, although necessary, mentor to Salcedo. “He brings nice power and thoughtfulness to his work, and to supporting others within the [music technology and computation graduate] program,” Egozy notes.
Salcedo additionally took a category with Egozy, 21M.385/21M.585/6.4450 (Interactive Music Techniques), which additional fed his urge for food for the creativity he craved whereas additionally permitting him to indulge his fascination with music’s prospects. By profiting from programs within the HASS curriculum, he additional developed his understanding of music principle and associated applied sciences.
“I took a category with professor Leslie Tilley, 21M.240 (Critically Considering in Music), which helped set up a beneficial framework for understanding music making,” he says, “whereas a category like 6.3000 (Sign Processing) helped me join instinct with science.”
Working throughout disciplines
Whereas Salcedo is keen about his music and his analysis, he’s additionally invested in constructing relationships along with his fellow college students. He’s a member of the fraternity Sigma Nu, the place he says he “discovered a house and group.” He additionally took a MISTI journey to Chile in summer season 2023, the place he carried out music know-how analysis. Salcedo praises the tradition of camaraderie at MIT and is grateful for its affect on his work as a scholar. “MIT has taught me the way to study,” he says.
Professors inspired him to current his analysis and findings. He offered his work — Synthetic Dancing Intelligence: Neural Mobile Automata for Visible Efficiency of Music — on the Affiliation for the Development of Synthetic Intelligence convention in Singapore in January 2026.
Salcedo believes his analysis can probably transfer past music visualization. “What if we may enhance the methods we mannequin self-organized programs?” he asks. “That’s, programs like multicellular organisms, flocks of birds, or societies that work together regionally however exhibit fascinating behaviors.” Any system, Salcedo says, the place the entire is greater than the sum of its components.
Growing the know-how used to design his software can probably assist reply necessary moral questions concerning AI’s continued enlargement and development. The trail to his work’s improvement is each daunting and lonely, however these challenges feed his work ethic.
“It’s intimidating to pursue this path when the academy is at present centered on LLMs,” he says. “However it’s additionally necessary to elucidate and discover the bottom know-how earlier than digging into extra nuanced work, which may help audiences perceive it higher.” Understanding that he has the help of his professors helps Salcedo preserve pleasure for his concepts. “They solely ask that we floor our pursuits in analysis,” he says.
His investigations are impacting his work as a musician. “My music has gotten extra fascinating due to the lessons I’m taking,” he says. He’s additionally all in favour of understanding whose music the academy and the world hears, exploring biases towards Western music within the canon and exploring the way to cut back biases associated to which sorts of music are valued.
“The work we do as technologists is way much less subjective than we’re led to imagine,” he believes.
Salcedo is particularly grateful for the help he’s acquired throughout his time at MIT. “Program college encourage a wide range of pursuits,” he says, “and ask us to advance our particular person goals slightly than specializing in theirs.” Throughout his time within the graduate program, he notes with enthusiasm how usually he’s been challenged to pursue his concepts.
Finally, Salcedo desires folks to expertise the enjoyment he feels working on the intersection of the humanities and the sciences. Music and know-how impression almost everybody. Inviting audiences into his laboratory as contributors within the artistic and analysis processes provides the identical form of satisfaction he will get from crafting a terrific beat or fixing for a thorny technical problem. Serving to audiences perceive his work’s worth fuels his drive to succeed.
“I would like customers to really feel motion and discover sounds and their impression extra totally,” he says.
