Artificial intelligence and the music school

One thing is clear: AI is raising eyebrows. Countless songs created by artificial intelligence are already circulating on the internet. Miley Cyrus sings Radiohead? No problem, and amazingly good. The music world is undoubtedly facing a wave whose effects cannot yet be predicted. The same also applies in the educational context, where the question arises as to whether an AI interlocutor could support pupils in learning an instrument well enough to at least partially replace the teacher. On the other hand, like the printing press and the internet, AI is not going to disappear. Riding the wave cautiously - to stay with the image - could also be worthwhile for instrumental and vocal teachers. AI can inspire, make processes more efficient, motivate students and relieve teachers. The tools listed below are currently free or cost between 20 and 40 francs per year.

Playbacks

AI can already save a lot of time in this area. One interesting example is Lalal.ai. This allows voices to be subtracted at the touch of a button - for practicing or playback of a cutting-edge song. All in one Moises. Songs can be divided into up to five individual tracks and extracted as desired. The tempos of the extracted tracks can then be adjusted and transposed to other keys.

Experimenting with songwriting, learning musical styles

With AIVA you can choose from a variety of styles and set the length, tempo and time signature. The tool then generates a song whose tracks can then be adjusted as required. Works in a similar way Boomywhere additional audio tracks can be recorded - including your own vocals. At present, the results of both tools are more likely to be classified as Saturday evening gimmicks. However, one of the presumed future market leaders MusicLM from Google is already in the starting blocks, and interested parties can currently join the waiting list. The direct competition MusicGen from Meta only creates short snippets of 12 seconds each at this point, but the first audio samples already show a big jump between the first and last two.

=> Listen to here first examples of

Music notation

ScoreCloud describes itself as "the most intelligent music notation software".  The tool creates editable scores from midi or audio files - if you like, directly from the vocals. The results are amazing. Students can quickly create their own songs at home.

Transcription

With Chordify you run the desired song and receive fingering charts for piano or guitar at the same time. This allows students to select songs for their own musical tastes at home and play along directly without the teacher having to search for and transcribe videos or audio files first.

Compile teaching materials

That ChatGPT is known to be quite liberal with facts. Nevertheless, it is (just like Bard from Google) is already very useful as a free version if you lack inspiration for a suitable piece for a specific student or a specific topic - or if you want to write letters, summarize long texts or put together a social media strategy. Jasper is subject to a charge, but is even more helpful because it provides links with further information.

Administration

Numerous mail programs already offer AI support. With Gmail, for example, "Smart Reply" is helpful: it prescribes friendly emails in different variants so that you can reply at the touch of a button. Another interesting feature is "Summary Cards", which presents important information from emails in a clear visual format. Something similar is offered by Spark Mail. AI will also be very useful for various everyday tasks, compiling timetables and taking notes from meetings - for example, the following is already available workstream.ai.

Feedback welcome! 

Do you already work regularly with AI? Have you put a particular tool through its paces? Or do you see other useful applications for AI in the day-to-day running of a music school? Get in touch with the editorial team and we will contact you for a short interview. Thank you in advance, we look forward to hearing from you!

From generative AI Dall-E generated images