With internet penetration, music digitisation has literally opened the flood gates. This is eventually having a repealing effect leading to more cultural divergence and inclusiveness.
The most powerful medium to evoke a spectrum of sentiments is music. Music has the dexterity to imbibe patriotism and uplift our spirit or make us dance and even bring back good old memories and emotions we may have forgotten we had, changing our mood within minutes.
Rhythms, pace, tempo, lyrics can actually transform us completely to different mood landscapes. Music is very personalised and all about feelings than anything else and thereby have immense psychological influence. With globalisation we are exposed to different civilisations, traditions and practices, which has broadened cultural acceptance and expanded our horizon towards music.
DIGITISATION OF MUSIC
With internet penetration, music digitisation has literally opened the flood gates. This is eventually having a repealing effect leading tomore cultural divergence and inclusiveness. We are seldom non plussed hearing tracks of a particular ethnicity in a different geography. We gyrate to bhangra numbers in Birmingham and hear Naatu Naatu at Oscars in Los Angles. Music Digitisation has completely transposed the music industry. It’s success is based on it’s win-win proposition:
• Artists now have extensive global reach.
• Online retailers are no longer constrained with shelf space and dead inventory that traditional brick-and-mortar stores were challenged with.
• Consumers have limitless access to music at a fraction of cost what they paid earlier, with the liberty to access any and all labels available across the world.
As on date we mostly enjoy our music through streaming services, with multiple players in the market viz. YouTube, iTunes, Spotify, Deeser providing music streaming service options. Covid was the turning point, global music subscriptions surged 26.4 per cent to 523.9 million during the pandemic, largely driven by Gen Z.
THE RISE OF THE ALGORITHMS
iTunes being the pioneer in digital music distribution. Spotify with half a billion users around the world, including almost 190 million paid subscribers, is the most popular streaming platform. Cross-platform streaming is also popular. For example, 42 per cent of Spotify users also use SoundCloud and 31 per cent also use Apple Music. With 15.03 per cent of CAGR (Compound Annual Growth Rate) in paid subscribers. Popular streaming platforms access large scale datasets and gain insight on global/local trends, then localise and publish playlists independently. By the way what is a playlist? It’s nothing beyond a library. We have playlists available for almost every occasion, be it different moods, activities, genres, eras, new music, top hits, emerging artists, cultural moments, regional listening trends and the list goes on.
Spotify already claims that one in every five Spotify subscriber streams attribute to its algorithmic recommendations. As on date a Playlist means accessing data hosted in the cloud and clubbing tracks with respect to genre, most played etc. and serving them. The Cache memory, between the CPU and main memory where the album art sits in our hardware, enables instant access to the Playlist. Today we often keep jumping from one playlist to another depending on time of the day, occasion or activity to induce specific emotions.
Having mentioned that, let’s not ignore the fact that we still create our own Playlists for varied occasions, which is unique to our predilection viz. for Birthday parties, Outdoor picnics, Independence Day celebrations, Religious gatherings etc. and the list can go on. It invariably implies loads of effort discovering new releases, patiently navigating through our memories and adding our favourite tracks inorder to create our playlist to be apt for a particular occasion. These tasks are often boring and time consuming. But the larger question is, why do we even do such futile exercises when streaming services already churn out playlists for almost every occasions? To answer that we need to understand the human mind i.e idiosyncratic, creative and rapidly evolving. Generic Playlists fail to meet our latent expectations.
We believe our personalities are build through our education background and professional experiences, however ignoring socio demographic factors viz. cultural background, social exposure, age etc. which have a massive impact on who we are and defines our preferences. Every individual has diverse sub-genre preferences which generally gravitates us to unique representations. Thereby categorising our musical taste into homogenous groups viz. genre or artist could be a presupposition.
PERSONALISATION IS THE KEY
Won’t it be amazing if someone could just map our different moods and seamlessly regurgitate tracks to keep our playlist fresh and happening, no matter what the occasion or mood is. That brings us to the question, how would an algorithm even understand our individual preferences and serve us musical tracks as per our personal choice? Specially, when it’s an uphill tasks even for us to construct the complex nuance and clearly describe our own personal taste for music. With technology it’s now possible to out source creating personalised playlists:
With data visualisation all the music tracks stored in our different Playlists are sliced based on information the track carries viz. language, genres and sub genre, artists, album, composer, date of release etc. and put into different buckets. ML (Machine Learning) then studies every bucket to identify analogy.
II. With IOT (Internet Of Things), AI (Artificial Intelligence) tracks our different sources viz PC, tablet, radio, mobile, watch, TV etc. that we use to enjoy music. May be radio while we drive to work and Spotify from our mobile phone when go for a jog. Further, what kind of music we enjoy in the morning when we go to the gym vis a vis when we party over the weekend with friends late evening.Thereby map our preferences based on location, time, frequency of repeat of track / artist etc. and create a pseudo personality of the individual.
III. Through every interaction with music, AI (Artificial Intelligence) begins to gradually define mood shades at an individual level based on time, activity, location and source.
IV. On top of that AI (Artificial Intelligence) would add another layer of personal information that comprises of age, ethnicity, work place/ residential address with respect to town/city, buying habits and draw out a clearer personality. Thereon correlate with music taste of millions of individuals with similar lifestyle.
V. Then AI (Artificial Intelligence) would correlate the current database existing in different playlist and apply all it’s learnings and serve new tracks, artists etc. as per the individual’s taste from the universe of music available globally. This could even include music tracks which are unheard of or not popular but matches our unique preferences.
VI. Depending on how often we react to recommendation of AI (Artificial Intelligence) in terms of appreciate or reject, AI (ArtificialIntelligence) will relentlessly keep learning and sharpening its search to personalise and recommend.
VII. With time AI (Artificial Intelligence) will be able to churn out personalised playlist for different occasions, which may even exceed our expectations.
Outsourcing music would make us indolent and eventually increase consumer dependability on AI (Artificial Intelligence). With change in order, various hygiene factors need to be aligned through policies that are agreed and honoured internationally. Checks and balances need to be agreed upon and designed in order to avoid promotion of hate or propaganda through music and songs. Which could disturb law and order. Artist revenue policies will also need to be aligned else record companies could exploit talent.
Fair competition clauses need to be inserted, to discourage large Record companies to create algorithms and insert various filters, leading to dominance, eventually squeezing out smaller players. The need of the hour is to create a forum of stalwarts across the entertainment industry, legal professionals and technocrats with representation from across the globe and draw out a frame work and gain buy-in from all stake holders.