Music historians Chris Smith and Tom Irvine are in their podcast bringing their points of view on environmental impact of the digital media and streaming services in comparison to other musical media of ‘post-consumption era’ as well as ‘pre-consumption era’. Those were marked by transition from purely acoustic music to electronically recorded and reproduced to recording media like shellac records, vinyls, CDs etc.

“Every system of inscription is tied to a system of extraction. Every discourse network is a resource network.” (Devine, 2019)

Chris’s and Tom’s main outcome, based on Kyle Devine’s conclusion, is that every single type of record media which emerged in capitalistic consumer society is somehow linked to extraction of natural resources as well as contributing to environmental problems as many other human activities. An interesting point is highlighting the fact that apparent current streaming services aren’t less damaging than for example CDs, vinyls or shellac records in the past. Behind every streaming service stand huge servers consuming enormous amounts of electric energy so they can continuously work.

They also came with in my opinion very interesting historical analogy comparing this invisible impact or toll to the invisibility of sugar cane business prosperity being inherently rooted and dependent on trans-Atlantic slave trade. There has been happening an apparent progress in British cities like Bristol or Southhampton in 18th century but not many people could actually see that this ‘progress’ has been built on exploitation and enslavement of a large part of humanity. Similarly people cannot see the invisible impact of many many data servers on the carbon imprint and the climate change.

Another interesting topic which Chris and Tom touched is AI generated music, particularly AI Jazz. Is AI Jazz problematic? They came to the conclusion that people who make AI algorithms don’t know that algorithms are working and ‘risk reducers’, in this case reducing risk of ‘wrong’ musical decision, thus they are influencing the composition into something which is less creative and basically getting stuck at the same sounding composition over and over again. The human element of improvisational base of the jazz encompasses remembering what other people had done and picking up what works for the musician and not what note to play at the certain time, which is what the machine does – quickly calculating musical decisions.

Personally I see a problems in other parts of AI music too. I am not fearing of musicians being replaced. AI can be utilised as a part of the performance and compositions but the fully generated AI music might mis-place the historical background of music genres or sort of erasing those backgrounds. This could be extended to the whole ‘genre’ or approach of AI music. AI algorithm doesn’t know where for example jazz comes from and doesn’t know its historical background.

https://www.soundinghistorypodcast.com/episodes/episode-8

Devine, K. (2019). Decomposed : the political ecology of music. Cambridge, Ma: The Mit Press.