Songwriters of global hits getting sued for alleged plagiarism has become a recurrent story on MBW these past few years – and a recurrent source of misery for writers and their representatives in the industry.
But what if a songwriter or composer were able to use AI technology to avoid litigation altogether, by finding out if their song copies elements of other compositions, potentially in real time?
According to a document published last week, Daniel Ek’s company is seeking a patent for its “Plagiarism Risk Detector And Interface” technology, which pertains to “Methods, systems and computer program products..for testing a lead sheet for plagiarism”.
As explained in the filing – and as our songwriter/musician readers will already know – a ‘lead sheet’ is a type of music score or musical notation for songs denoting their melody, chords and sometimes lyrics or additional notes.
Spotify’s invention would allow for a lead sheet to be fed through the platform’s ‘plagiarism detector’, which would then, “having been trained on a plurality of preexisting encoded lead sheets”, immediately compare the composition in question to all other songs stored in its database.
A set of messages would then be displayed – describing a detected level of plagiarism regarding “a plurality of elements” such as a chord sequence, melodic fragments, harmony, etc. of a song (see fig 7 below).
The AI software would also potentially calculate “a similarity value” of the song in question vs. other songs in the Spotify lead sheet library.
These technology could work the other way around, too, says Spotify’s filing, reassuring a songwriter that “the melodic fragment [of your song] appears to be completely new”.
One particularly interesting element of this is that it would take place in near-real time, allowing a songwriter or composer to tweak elements of their work to avoid infringement before they (and/or their record label) spent the big bucks on recording a final version.
Spotify’s filing adds that “in some embodiments a link to the media content item that might be infringed (e.g., a track of an album) is provided so that a [songwriter] can quickly… listen to the potentially plagiarized work”.
MPs are to investigate the impact that music streaming is having on artists, record labels and the sustainability of the music industry. The Commons digital, culture, media and sport select committee will examine the business models of major streaming companies such as Spotify, Apple Music, Amazon Music and Google Play to determine their fairness towards the writers and performers whose material populates the platforms.
Committee chair, MP Julian Knight, said it would look at whether “the economics of streaming could in future limit the range of artists and music that we’re all able to enjoy today”, describing the algorithms that promote music discovery on these platforms as a “blunt tool to operate in a creative industry with emerging talent risking failing the first hurdle”.
Music streaming in the UK brings in more than £1bn a year in revenue. “However, artists can be paid as little as 13% of the income generated,” a statement said. Spotify is thought to pay between £0.002 and £0.0038 per stream. Apple Music pays about £0.0059.
Spotify is testing what it describes as a new “listening experience” that combines music and spoken-word content in a new playlist-style podcast format.
Shows that use this new format, which the company says is one of its most-requested features, will see talk segments and individual tracks stacked together, allowing Spotify’s listeners to interact with music within episodes in the same way they normally interact with tracks on the platform.
The company states that this new format marks an evolution from its personalized Your Daily Drive podcasts that combine podcasts and music, which were launched in the US in June last year, followed by Germany in October 2019 and in the UK last month.
Shows using Spotify’s new format rely on the platform’s music catalog licenses, which means that musicians and songwriters are compensated “just like any regular stream of a music track on Spotify”.
Spotify Premium listeners will hear full tracks as part of these shows, and Free tier users will hear 30-second music previews.
Shows using this format are also exclusive to Spotify.
Spotify has been granted a flurry of US patents in recent months.
MBW discovered in September, for example, that the platform had been granted a patent for a new karaoke-like feature that allows users to “overlay a music track with their own vocals”.
And last week, we told you about a new Spotify patent that suggests the company is working on geo-targeted advertising, using 3D audio.
Today, MBW’s gotten hold of a new Spotify patent. It may well prove to be transformational for the platform – but it’s also, and there’s really no other way to put this, really quite creepy.
Yesterday (October 6), Spotify was granted a US patent for “Methods and systems for personalizing user experience based on [user] personality traits.”
The patent application was initially filed in November 2018.
The filing, leaked to us by a source close to SPOT argues that “there is a need for systems and methods for personalizing media content in accordance with one or more personality traits associated with a user”.
Behavioural variables such as a user’s mood, their favourite genre of music, or their demographic can all “correspond to different personality traits of a user”, reads the patent.
“Thus, it is possible to identify a personality trait of a user based on the content (e.g., music) the user consumes (e.g., listens to) and the context in which they consume the content.”
Spotify claims that “before assigning a personality trait to the user, a personality model may be built”.
The patent clarifies that this “model” could identify users’ personality traits “based on a questionnaire, such as the Big Five Inventory (BFI-44) or the Meyers-Briggs personality survey“.
“The traits measured by the questionnaire may then be used as the possible personality traits that can be assigned to the user. [These] possible personality traits… include the Big Five personality traits: openness, conscientiousness, extraversion, agreeableness, and neuroticism.”
Spotify suggests that, armed with this knowledge, it could then promote content – presumably audio advertising content, but also perhaps music and podcast content – to users based on the personality traits it has detected.
This is where things starts getting a little that-moment-when-things-turn-bad-in-sci-fi-movie.
“In some embodiments in which the personalized content includes one or more messages with audio components,” explains the patent, “the electronic device changes a tone of voice for messages for presentation to the user.”
It continues: “For example, the tone of voice may be more upbeat, high-pitched and/or exciting for users that have been assigned the personality trait of extroversion.
“The tone of voice may be quiet and/or soft-toned for users that have been assigned the personality trait of introversion. This modulation of tone helps to humanize the user interface for the media-providing service in accordance with the user’s personality, thereby improving the user experience.”
The science behind the filing is more than a little unnerving, too.
Several individuals named as inventors of Spotify’s patent – including Ian Anderson (A Senior Research Scientist at Spotify), Clay Gibson (Senior Machine Learning Engineer at Spotify), Scott Wolf (a Data Scientist at Spotify) – co-wrote a scientific research article published in July this year.
Entitled, “Just the Way You Are”: Linking Music Listening on Spotify and Personality, the article details a research project in which the authors analysed 17.6 million songs and over 662,000 hours of music listened to by 5,808 Spotify users in the United States over a three-month period.
The summary of the article explains: “Building on interactionist theories, we investigated the link between personality traits and music listening behavior, described by an extensive set of 211 mood, genre, demographic, and behavioral metrics.
“Findings from machine learning showed that the Big Five personality traits are predicted by musical preferences and habitual listening behaviors with moderate to high accuracy.”
The study was given ethical approval by the internal review board (IRB) at Spotify, according to the article.
Its authors conclude: “The present work is a model for how psychological methods can be fused with cutting-edge technology and big data for scientific inquiry.
“[The] results show that personality does in indeed play an important role in musical preferences and warrants continued rigorous investigation.”
What might this “continued rigorous investigation” comprise, exactly?
From the article: “Given the vast volume of research on the cognitive neuroscience of music and the emerging literature (Peretz & Zatorre, 2012) on the social neuroscience of music (e.g., the role of oxytocin) (Keeler et al., 2015) , future research could begin to link streaming behavior with brain scanning, genetic, and physiological data.”
(Oxytocin, if you weren’t aware, is a hormone secreted by the body’s pituitary gland. Coined the “love hormone”, it is typically released within humans when a couple are sharing a tactile moment – e.g. cuddling on a sofa – or during a strong social bonding moment.)
The Spotify article’s authors warn that “such future research and applications must be conducted within the strict boundaries of ethical data usage, collection, and storage policies”.
They add: “A user’s digital history is extraordinarily personal and sensitive and should be treated with proper consideration of the conceivable misuses and unintended externalities.”
We like reading that bit in our head in the voice of Dr. John Hammond, from Jurassic Park.
Last year, the journalist Liz Pelly wrote an article called “Big Mood Machine” in which she looked at Spotify’s existing mood-related data analysis for content recommendation as well as advertising purposes.
Pelly’s closing observations in that article serve as a profoundly relevant analysis of Spotify’s latest patent.
“We should admit that it’s good for business for Spotify to manipulate people’s emotions on the platform toward feelings of chillness, contentment, and happiness,” wrote Pelly.
“This has immense consequences for music, of course, but what does it mean for news and politics and culture at large, as the platform is set to play a bigger role in mediating all of the above, especially as its podcasting efforts grow?”
With the Music industry at a stand still with numerous festival and artist live shows being cancelled indefinitely or postponed to next year. With convid – 19 affecting the world, streaming, and digital downloads become all the more important to indie artists. Spotify have acted by introducing a ‘Tipping’ Service to the platform, where fans can donate either to worthy causes selected by the artist or to the artist themselves via PayPal.me and Cash App.
Spotify said in a blog post: “Given the urgency and impact of the COVID-19 crisis, we’re working as quickly as we can to develop this new product and get it out to as many artists as possible. However, we’ve never built a fundraising feature like this before. We consider this a first version that will evolve as we learn how to make it as helpful as possible for the music community.”
If this feature were kept permanently, we could see it having a huge impact to indie artists’ in particular in how much smaller artist get to the platform could have to smaller artists in particular.
Jazz pianist and producer, Jake Milliner explains the potential impact on Instagram: