This may be true, but can a modern machine or AI, on its own, give a detailed and personalized explanation of why the student was incorrect? For instance, in terms of music, I do not believe a machine can explain the nuance and tone of a passage. It may be able to play a professional recording, but in my opinion, music-making, especially at an enriching, educational level, should be a creative process, not a reductive, emulative one.

Furthermore, there is the problem of the expenses associated with these technologies. Let's say, in 2019, a machine or software is created that can grade music theory assignments with 99% accuracy. How long would it actually take for a significant number of schools to adopt such an AI? While wondering how great it would be to have such a device, it is simply not useful to pretend that it is already here.

Beyond Scantron multiple choice graders or online assignments or videos, I rarely see machines that take the teacher's role. No machine could do everything a human teacher does in this day and age.

Perhaps I extrapolated too much from this article. However, in my mind, when I see someone talk about "machine learning" or "machine teaching," I think of neural networking, big data, and Google Deep Mind.

To create a music-teaching machine, expanding upon my first annotation, one would need it to understand the musical material. We would need complex, large-scale neural networking that can compare a student's playing to some model or professionally-done recording. With regard to my own philosophy, I think this would lead to a lack of uniqueness among young musicians. Additionally, this software would need to be easy for teachers to use without programming experience.

In short, if we are to use machines to their full, modern capacity to inspire and guide young musicians, we would need:

1. Neural networking software,
2. A simple-to-use way for teachers to access or edit that software,
3. recording equipment.

Perhaps I extrapolated too much from this article. However, in my mind, when I see someone talk about "machine learning" or "machine teaching," I think of neural networking, big data, and Google Deep Mind.

I see positive reinforcement very often. The directors often tell students when they are doing a good job, delegate solos to young artists, commend the ensemble for their work at the concert in front of the audience, and sometimes even let the band decide how a passage should be played. Much of a band, choir, or orchestra director's job is to give positive reinforcement. I'd even say that receiving this praise is the goal of some students for one reason or another.

Some of these examples can only be entertained in a music classroom fit for them. For instance, in an ensemble where the semester's concert is the final, this would not work. In other contexts, such as a general music, music theory, or music appreciation environment (or even ensembles that do have separate exams), I could see an optional exam, dropped poorest assignment, or homework pass being realistic.

My point? I rarely see negative reinforcement in music classrooms; when I do see it, it serves a very specific purpose. While negative reinforcement is not common, it is useful, and I'd like to see good examples of it applied to a large classroom or ensemble rehearsal.

It appear to me that as of 2018, Jacobs is ahead of this curve; to me, this curve has passed in teacher education. In terms of technology and new course material, I know we have had music education lectures and courses in the past on iPad ensembles, basic guitar performance, body percussion, and the pedagogy of pop music. In some other courses, we have networks of electric keyboards the instructor can listen to from their seat, and the Department of Bands has taken advantage of projectors and movies just like some major symphonies do.

I'd say that this misconception can be falsified by simply stepping into a real teacher's classroom. I believe that there is a difference between pedagogy and execution, or an idea and its reality. I have had several of my own teachers fail to attempt this misguided approach with the students just more confused after the teacher prompts them. In all these cases, the instructor meant well and obviously intended for us to reach a certain conclusion, but nobody was able to get there on their own without direct instruction.

In a nutshell, this section provides an example of a subject that relies on intuitive thinking or additional knowledge to fill in the gaps between its parts. Knowledge does not arise in the form of lists; it could be said to be more like a web, or a mental internet of ideas.

Annotating is cool