Guest post by Natalie Tsaldarakis
I came across a video of an exasperated ballet dancer talking about parents who had turned to AI to correct their children’s posture. It prompted this article, a response to the dizzying speed at which AI is infiltrating every facet of our lives. The uptake is so fast that the night before, I had been reading through people’s responses to a question about jobs they thought were safe from AI-driven redundancy. Several suggested cleaning or other manual work, apparently unaware of another post showing a robot slowly but methodically cleaning a house and neatly folding clothes. One thing is for certain: we still can’t grasp how our lives will change in the next few months and years.
Meanwhile, my path as pianist, teacher, musicologist has had a surprising turn: I am conducting a pilot study with Imperial College, on the biomechanics of elite piano performance and training, with access to resources provided by Steinway Hall and Goldsmiths College’s Deptford Town Hall. AI is not going to trail very far behind: there is 3D body imaging, motion tracking and data to analyse. In this narrower respect, AI is simply an augmentation of a computer’s abilities.
Since the advent of the internet, availability of information masks from view the more concerning shrinking of the new ‘reality’ into a colourful yet ‘flat’ experience: that of the phone, tablets and tv screens. Everything is one, or a few swipes away. In the case of a non-domain expert parent, i.e. someone who hasn’t trained in the same subject, AI can provide some background information fast, provided the right questions are asked and provided that the AI’s limitations are acknowledged. AI itself is not a ‘knowing’ entity, but a processing entity. The instigator and recipient of this knowledge is the human. A deceptive communication, which is actually cyclical: it reflects and engages with the human input in a seeming loop. So, yes, the information is available, but the context, in terms of indispensable layers such as aesthetic pursuits and creativity, is missing.
If we try to probe the matter deeper, AIs such as the general public uses, Gemini, Claude, ChatGPT, are actually language models. The basis of their functionality is that of predicting a probability distribution over next tokens and sampling from it to arrive at cogent and context-realistic verbal responses. And I am reminded of a very pertinent excerpt highlighting the process in humans, which connects internal thought processes with language, from Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf, edited and with an
introduction by John B. Carroll, Foreword by Stuart Chase
The Greeks took it for granted that back of language [sic] was a universal, uncontaminated essence of reason, shared by all men, at least by all thinkers. […] Whorf flatly challenges it in his second major hypothesis. “A change in language,” he says, “can transform our appreciation of the Cosmos.”
Taken simply, internal processes of thought and experience go at least hand-in-hand with language. Although AI are language models, their output is not an interaction in the sense of having the initiative to communicate unprompted or the result of thought: the initial stimulus is the user’s thought and it is anchored in that. It could be that our over-reliance on AI ‘transforms our appreciation of the Cosmos’ in unintended ways. This is certainly
unchartered territory. What about the teacher’s concern over AI teaching movement without human supervision? Expert skill acquisition, such as required for elite ballet dancing or piano performance, is narrow and deep. The idea that anyone will do because ‘the child will not pursue seriously music’ is wrong on so many levels it requires a separate article. And I think that it is this attitude which operates in the background when parents think that AI can be a good substitute for a teacher or a sufficient supplement. The generalisations that come out of Google currently have to do both with data poverty and the fact that biomechanics, the abstraction of art to physical movement, are the necessary substrate of what we call Art, but these two cannot be equated. A good teacher is simply indispensable to instil the basics for specific outcomes to be achieved. Strategy is also not the strong point of AIs.
Teaching is an interaction, so the process is not just a passing down or the rehearsing of a physical skill, but a process within a social context. For non-expert parents, the necessary critical faculty to efficiently sift through sometimes conflicting information is lacking precisely because the requisite is not just information; it is lived experience. In her book Producing Excellence: The Making of Virtuosos, Izabela Wagner writes about this very idea
in terms of “socialisation” in the arts: you need a human to transmit not only physical technique but also the lived experience of the milieu you are trying to become a member of.
The Arts are quintessentially the means of probing the eternal questions that torture us, to echo nineteenth century sentiments, a time when many of the canonic works of classical music were composed and are heard regularly on concert stages around the world to this day. More to the point, unlike science which seeks to prove and to know, the Arts seek to engage, no matter what the answer might be, or even when the ‘answer’ represents, like Edvard Munch’s painting The Scream (1893), an expression that cannot be ‘answered’. It just is. Uniqueness would be the other identifier, discussed in connection to seemingly artless or mundane, now iconic, works such as Andy Warhol’s Campbell’s Soup Cans (1962): the ability to reproduce it, underlines its emphasis away from the skill of the craft and on the effect on its unsuspecting audiences. This then supplies the third aspect of human artistic production, the intent to communicate and the decision to take action towards this direction. This is what parents of young learners miss: to develop a real relationship with the arts is to engage with socialisation which cannot be abstracted or equated to ‘interaction’ with a machine and ultimately to connect through one of the most fundamental of human needs: love, and belonging.
1 Benjamin L. Whorf, Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf, edited and with
an introduction by John B. Carroll, Foreword by Stuart Chase (The Technology Press of Massachusetts Institute
of Technology and John Wiley & Sons, Inc. New York London, 1956) Foreword, p. 8.
Natalie Tsaldarakis is a Greek concert pianist, researcher, and co-Artistic Director of the PianoPlus International Recital Series, now in its 4th edition. Her research sits at the intersection of biomechanics and piano pedagogy, currently underway in collaboration with the MSk Lab at Imperial College London, and she is completing a forthcoming Routledge publication on the pianist and pedagogue Gordon Green (1905–81). A graduate of Royal Holloway, where her teachers included Chopin scholar Prof. John Rink, she performs internationally as a member of the Ivory Duo Piano
Ensemble and Convivium Label Artist, and is a published author (Classical Music in a Changing World, eds. Kramer & Nones, Vernon Press, 2021).


