Welcome to my personal website! I’m Victor Zhou, a researcher and musician with a deep passion for exploring the nature of the world I live in and understanding me myself as a human being.

I got my Bachelor of Science degree in physics in Fudan University, Shanghai as an outstanding graduate of the city. Beyond physics, I was a tenor of the university choir (FDUC) and got a national gold prize with them, passed both ABRSM's piano performance and music theory level 8 (the highest level). I was a winger of the University's handball team and we won 2 Shanghai Championship in 3 years and got the 8th place in the national tournament. I took various classes in subjects like economy, logic, history, chemistry, biology, psycology, philosophy, linguistics and I also gain sound trainings in programming in python, C and mathematica. In retrospect, the extensive and intensive learning during my undergraduate study laid the foundation of my interdisciplinary learning and collaboration later on.

I was seriously considering furthering my study in a conservatory later in my life so I went to exchange to the department of music in University of North Carolina, Greensboro (UNCG) for a semester during my junior year. Although this taste of the conservatory life confirmed my interest and capability to study music professionally, ironically, it also ignited my great interest for theoretical physics. The only physics-related course in UNCG is astrophysics (because UNCG has a very good observatory in the forest near the university, and it was really FUN!!!), but I spent a lot of time on reading different physics textbooks I borrowed from the library because I had more free time on the physics side. It is VERY DIFFERENT to learn something in chunk and having the time to delve deep (like refering to other materials on the same topic) when not understanding: I clearly felt that every course was trying to tell one single story and all these stories were contexts for the next. Although I didn't know whether I was talented enough to being able to contribute significantly towards the Theory of Everything, I definitely wanted to learn the existing and established theories of our world and prepare myself so that if later some breakthrough happens in our understanding of the world, I would be able to understand it first-handedly through the original article instead of from easier-to-understand-but-mostly-wrong popular magazines.

So I went on to pursue the master of physics (theory track) in University of Amsterdam where I delved myself into learning the two pillars of the cutting edge quantum gravity: Quantum Field Theory (QFT) and General Relativity (GR). These two theories are so refreshing and transformative that I actually encourage anyone who have time and energy to learn because it will definitely worth your time (If you need advice or suggestions regarding the textbooks or methods/roadmaps in learning these, feel free to reach out! I may write an introduction later myself. The best textbook for GR is definitely Differential Geometry and General Relativity by Canbing Liang, but that for QFT will have different choices depending on your learning style and the specific topics).

I love Amsterdam for various reasons, the clean city, the canels (I love rowing and boating), the bikes, the flowers, the oranges (yes! but only from AH), the RCO, the museums, and the fantastic collision of modern and classics (take the ferry, try the modernity of the North and the classics in the South). But unfortunately, I arrived in Amsterdam during the fierce attack of Corona, school closed, restaurants and museums closed, and there were even a curfew out there. That adding to the fact that in winter, it was completely dark around 4:30pm and the road lamp was still on until 9am, makes me completely understand Vincent van Gogh's early style of painting (very DARK). It was then that I started seriously singer songwriting and published quite a bit of them to online platforms like QQ music and NetEase Cloud Music. The songs were described as "sentimental", possibily because the weather and Corona then.

I got my Master's degree of physics from University of Amsterdam and was admitted to the Master of Music program of piano performance major in San Francisco Conservatory of Music in the joint studio of Prof. Corey McVicar and Prof. Yoshi Nagai. I decided to do a second master in music for various reasons.

First, before the Corona, the first chairs of violin and cello of Shanghai Student Symphony invite me to perform the Mendelssohn's first piano trio in D minor with them. That was my first chamber performance experience and it was mind-blowing. I really craved for other opportunities to collaborate with great musicians and conservatory can give me the chance.

Second, when doing mixing and recording jobs, although physics background provided me the ease at handling the engineering and theory details, I still need to test and compare different equipments and plugins using my ear (ultimately it is ear training) but good microphones, speakers or plugins were notoriously expensive and the conservatory could grand me an opportunity to try out a vast amount of industrial standards and come up with my own favorites. SFCM's TAC (Technological and Applied Composition) major has a Grammy-Level recording and listening studio (where later I did the advance recording class and various recording sessions) so that's a huge plus.

Third, I had difficulties in composing music using instruments other than piano, basic strings and voice. I didn't have the opportunity to know anyone playing the brass, woodwind instruments, or percussions (and harp) so it is hard to write in these instruments. The Royal Concertgebouw Orchestra (RCO) is one of the best orchestras in the world and the student tickets are very inexpensive so I went there almost every week. This exposure gave me the taste for good symphonic music and when I wrote music, a lot of times the symphonic sound will emerge in my mind but it didn't give me enough material to learn properly how to write them (sometimes I could feel the correct instrumentation and successfully implement them out but most of the times it was very hard). I thought I could learn a lot from composers and musicians playing other instruments in the orchestra in the conservatory.

The last reason is of course, piano performance itself. I could feel the big gap between my piano playing and the professional level but didn't know how to close it so I want to learn from professionals and fellow musicians and polish my skills in various performances.

Just as I have expected, the two years in the conservatory was very intense but fruitful. I got every thing I want and even more. I collaborated in chamber music a lot, with including violin, cello, viola, double bass, flute, oboe, clarinet, bass clarinet, bassoon, french horn, voice, harp and trombone (I tried to collaborate with all orchestral instruments indeed). I played a piano concerto for the first time, recorded my compositions in professional studios, premiered my own compositions live, premiered other composer's work in various concert halls, had two personal recitals both of which last over one hour, performed in a lot of different venues around the bay area, went to music festivals for the first time and performed in a church at the coast of Italy.

During these two years, I also passed CFA level I and learned Coursera courses on Machine Learning, NLP and Java programming. The reason of my studying the CFA was because my ignorance of financial and accounting knowledges started to annoy me when I need to manage my own finance. And the reason of studying AI related courses are of course, because of ChatGPT. I have been exposed to chatbots and natural language processing before in my undergraduate study when I learned basic linguistics but I had never imagined the NLP can achieve this kind of intelligence: solely based on "interpolation and regression". Another shock to me was the Ultimate Vocal Remover (UVR) which incorporates several cutting edge AI model for Music Source Separation (MSS). I have long been using MSS to help me analyze pop musics and produce remixes, but most of the classical algorithms were very basic and not performing well. The AI algorithms (for example the Demucs), on the other hand, provided fantastic effects on separation (at least among the pop band instruments and vocals), which has already benefits me and my friends a lot when practicing writing background musics to films: previously if we want to practice this, we pack the video with our musics, which, inevitably, missed any dialogues. Now, with the help of AI, we were able to separate the voice track with the instrumental track and substitute the origianl soundtrack with our music, this creates a much better "playback" for us to test "how well we wrote". Another "playback" application of AI is the Noteperformer which is a plugin on notation software like Sibelius: it studies the relationship of notes and fine tune the inner dynamic of a note so that the playback is more realistic. The Noteperformer is not open-sourced but the Demucs is, so I am currently using Demucs as basics and try out the MSS project on classical ensemble recordings.

I am interested in AI because of reasons in education too. As a lover in learning (and a lover in teaching, frankly speaking, based on Feynmann's method in learning, I do think this two aspects are non-separable), ChatGPT has been helping me greatly and I can see even bigger potential in it. Typically when I learned physics before, a common practice was to consult a lot of different textbooks because I need to find the one that resonate with my current knowledge and understanding more, sometimes it was very hard to do so. ChatGPT provides a distinctively different way of learning. Even if the debatable argument is true that the current LLM cannot produce new ideas and create truely new things, the flexibility in re-organizing the knowledge itself can be very beneficial in helping people learn complex, abstract or counter-intuitive ideas. A chinese saying goes "teaching students in accordance with their aptitude", this is now shown potential to be accessed by anyone.

I am very interested in this topic and want to know answers to questions like: How do people learn? If someone's prior knowledge and favorite working style is known and for a specific knowledge point, what will be the best roadmap to get to there for this person? What will be the systematic way to re-organize the knowledge such that it is most understandable for them? How does illustration or demonstration work? How to incorporate different sensory of human to better teach a topic (so-called embodiment cognition)? The answers might not be straightforward, as I have experienced in both myself and teaching, sometimes leaving certain areas blank (actually, controversially speaking, sometimes when the teacher made a mistake, I would learn better) will turn out to be beneficial for certain people because a doubt may arise and the curiosity will drive them in exploring the topic and finally achieve better understandings. Towards this, I am reading related cognitive science books for the human aspects and doing AI projects to be more familiar with AI project development and the logic behind AI.

Similar to this education application, another human-AI interaction that I am very interested in is that of music collaborations. This stems from my passion in chamber music: collaborating with other musicians of one piece is very different from playing by oneself and it is more than fulfilling when everything is in sync and your ideas can be captured by other musicians (even if that's because you have agreed upon it during rehearsals). These experience are currently limited to conservatory students. If the MSS of the classical ensemble comes through, it may unlock a broad dataset of past recordings of how musicians do interact/collaborate on stage via their respective separated tracks. We also have the Noteperformer kind of AI generation which may aid in the process. Together with some cognitive science and some rehearsals to learn the players style, I do think AI has the ability to make this collaborative music enjoying experience more accessible (enjoying music by listening alone and with yourself playing in it is very different! That's also part of the realm of embodiment cognition).

I am currently based in the San Francisco Bay Area. In my time additional to my performances, teaching, composing, learning and researches, I also like cooking, hiking and badminton. I recently found a very nice (free) archery range near Saratoga so I am expecting to pick my archery up too some times.

Thank you for visiting my page—feel free to reach out to collaborate, connect, or explore ideas together!