This week, the Economist has published a little article about a new technology allowing the electronic recognition and identification of music. It’s well worth reading, although the article does not hint at the enormous importance of this technology (which is still being developed, to be clear about this).
You will probably remember that the music industry is currently involved in a severe battle concerning the perpetuation of non-digital age copyright structures into the digital age. I have already explained at lenghth in this „diary“ why their success would be economically inefficient and socially problematic, to say the least.
What you haven’t heard a lot about in the last years is that the copyright battle is only one front of the war that is changing this world’s cultural landscape. And while you might think that the music recognition technology alluded to in article linked to above is only about a giving you a hand in remembering tunes you heard on the radio once or twice, its implications are much bigger. Think about it – most of the songs you hear on the radio today have been selected or produced by the oligopolic major labels for one reason or another. Given the enormous amount of music out there, we need some sort of screening as we do not have the time to listen to all of the material on the market – of which, in addition, we would not like a lot.
While screening the market is a necessary element in the „music value chain“, the gatekeeper function of the major record labels is somewhat problematic in a cultural sense. There’s a lot of music out there which you might like, but will never be able to hear because it lacks the commifying support of the record industry, aka marketing and distribution. Digital distribution has already reduced the importance of the latter element in that equation, but the first one is still the most important service provided by record labels. In fact, their basic role in the industry is that of a specialised venture capital provider – and given the traditional (non-digital) cost structure of marketing a new act and their average success/failure rates, the comparison does indeed make sense. As a consequence, our record stores are mirroring a „winner take all“-market, in which very few earn very much and most earn very little.
The internet, especially sites like, most prominently, mp3.com, now owned by the French media conglomerate Vivendi, have alredy reduced the salience of physical distribution for new acts. But even with advanced systems of collaborative filtering, comparing your personal taste to that of other users on the same system, a statistically significant amount of initial consumption is needed. So marketing is still an issue, although digital distribution has increased the options for a lot of musicians previously unable to earn a living by making music. However, for broad consumption, these days, they still need a bank.
But with advent of machine listening, as described above, things will probably change considerably. It is advancing a reduction in the expected value of a specific piece of music (now that value increases with the amount of marketing put behind an act) and thus creating a musical landscape less characterised by „winners who take all“. It is doing this by really cutting the middle man, the oligopolic gatekeeper’s and fund provider’s current commidification services, out of the connection between the artist and the consumer. Think of a big market place which artists can put their music on, and consumers can look for music they like, regardless of the support they have previously received from a label. The computer has a list of criteria to listen to, possibly including production quality, lyric content, and the like – I doubt, the support by record labels will be an important element of the song’s meta data.
No wonder the middle man is scared, don’t you think?