Letter 11 - Mission Statement

Virtuous curation, algorithmic playlists and independent radio.

Hi,

Whilst we all recover from the Christmas break, I recover from my now annual tradition of getting early January surgery, and as there’s a partial lull in new releases, I have been mulling over something I’ve been asked a lot in recent weeks - why on earth am I spending so much time, often very late at night, writing this newsletter.

So I’ve tried to answer this question, and I’ve recommended some of my favourite radio shows you should check out.

If you just want some music and film recommendations, jump straight to that radio section or have a look at my Favourite Albums and Songs of 2024 and my Top 25 (26…) Films of 2024. All of my other letters are accessible in the archive here.

Please like & subscribe, share with anyone who would enjoy, and let me know what you think.

Much love, Odhran x

An Answer in Two Parts:

Part 1 - Music and Film as Artforms

In my first letter, I said the idea behind The Odhracle was to recommend music and movies to my friends in one place, instead of the scattergun approach I’d adopted for years prior, but I wasn’t clear why I have always spent so much time doing this.

I considered whether it’s from a misplaced sense of superior taste, but I don’t (in the main) actually think I have that. Taste is subjective - I love films that many consider trash (Michael Bay’s Ambulance (2022)), hate some which are canonically classics (Before Sunrise (1995)), and I don’t actually think my favourite band is better than your favourite band (although I’ll admit I am sneering about some mainstream behemoths). I also wondered whether I do this because of some neurodivergent compulsion, but I have plenty of other obsessive hobbies and I don’t force them upon friends and family.

The reason I enjoy telling people what to watch or listen to is because I think engaging with music and film as art forms in a meaningful way is a critically important aspect of life, and in particular discovering music and film that’s not already within the confines of your pre-existing taste, and most people don’t do enough of it.

This mindset of course applies to other art forms but music and film are the art forms I know the most about, so there you go.

However, that’s only half an answer really - it doesn’t really explain why I think writing The Odhracle is a worthwhile exercise (other than as an outlet for the nonsense in my head) - what problem do I think I can help solve?

---

Part 2 - Finding music in the age of streaming

From this point on I’m mostly going to talk about music, as the problems I discuss mainly effect music and how we listen to it, but there are analogous issues in film (including TV) that I’ll discuss some other time.

In some ways, discovering music has never been easier with the advent of on-demand streaming, but with each passing year I am increasingly dissatisfied by what music the tech giants decide I want to consume as “content”. I see this dissatisfaction in others too, even if they don’t feel it as viscerally - I’ll ask friends if they’ve listened to anything new, and they’ll name something from their Spotify Discover Weekly or that they found on social media, in either case with any great enthusiasm.

In the early days of digital streaming, the average person was immediately exposed to vastly more music than ever before - millions of artists from across the world at our collective fingertips. But between content being removed from streaming platforms at the whims of rightsholders, consumers required to pay for several services to only have access to a fraction of content that could be available, and the ever-increasing prolificacy of AI and algorithm-led content selection, those halcyon days feel like a long time ago, and the streaming landscape is now a maze of smoke and mirrors which actually acts antagonistically to the goal of finding new music.

It’s not only technology that makes finding new music a challenge - nature plays a role too. Informal analysis of Spotify usage data indicates that “for the average listener, by their mid-30s, their tastes have matured, and they are who they are going to be” - in other words, by their mid-30s the average person stops listening to new music.

John Waters once said “As soon as you stop listening to new music, your life is over”, and I think that giving in to the apathy that stops you discovering new music as you get older is a choice, not a biological certainty, and we can and should resist. New music doesn’t have to mean recently released music which is popular with teenagers - I’m not asking you to listen to 100 gecs - but it can be the exploration of genres or artists that are new to you, even if released 20 years before you were born.

So simply put my goal is, in a small way, to try to help people to listen to more music, and in particular music that they otherwise wouldn’t be exposed to, because I think the landscape in which people can do so is becoming hostile to achieving that goal.

If you’re interested in finding out how the digital landscape makes discovering and meaningfully engaging with music more difficult, read on.

---

Music Is Art, Not Content

I’m primarily talking about Spotify but a lot of these issues apply in equal measures to other major streaming platforms. There’s also various criticisms of Spotify that I’m not discussing in this letter, such as paying artists an estimated average of $0.003 - $0.005 per stream and not paying artists anything for tracks that receive less than 100 streams.

A lot of the issues I discuss below adversely impact artists financially, making it harder for them to make music. I’ve not delved into that, but it’s why my newsletters link to artist’s Bandcamp pages, because it’s through buying music from the artists you like, going to a gig or buying some merch that you actually support them, rather than just hitting play on a streaming service and leaving it at that.

---

Active v passive listening

Prior to the advent of digital streaming, our engagement with music was mostly a more purposeful and focussed affair. You intentionally selected a CD, usually an album, put it in your player and pressed play. Once it finished, it either looped, you restarted it or you selected a different album and repeated the process. You listened to each song within the wider context of the album, and in the manner envisaged when it was recorded. The album would have been crafted with that experience in mind - telling you a story or conveying a feeling, a time or a place, as you listened to the album from start to finish.

Even if you listened to a mixtape (or in my case, a burned CD), it would have a limited number of songs on it, each carefully selected by you or someone for you, to be listened to in a specific order - a process which bestowed a significance and meaning.

For over a decade now however, the playlist has been king. I’m someone who has sprawling 50-hour long playlists, that I regularly hit shuffle on, and I know people who almost exclusively listen to music via their “Liked songs” playlist. There’s no inherent problem with this, it’s incredibly handy to keep a track of all the music I like, for it to be sorted into moods, settings and in specific company. You couldn’t tear my playlists out of my cold dead hands.

There is an inherent problem however, if this is the primary way in which you engage with music. Aside from anecdotal evidence that steaming platforms are incapable or unwilling to truly random shuffle longer playlists (hence why you think the same songs get played all the time on shuffle) - hitting shuffle and not paying attention to what’s playing next is a passive listening experience. You’re not listening to the music in context or engaging with why it was written - it’s a more superficial experience.

This passivity, or as Spotify calls it “lean back listening”, is a stated goal of the streaming giant. Streaming companies want you to listen constantly, all the time, in the background. Constant background listening is more minutes on the clock, and an instinctive thing that users do. It creates an environment where users are less likely to switch off or unsubscribe. This becomes problematic when it becomes the primary way people listen to music.

This shift towards passive listening has had a demonstrable impact on how music is recorded, with less emphasis placed on the album as a creative concept, and more on singular songs that may work their way into curated or algorithm-led playlists. Spotify itself eschews the album as a concept, not including albums in their 2024 edition of "Wrapped”, and in contrast to other streaming services, not including any album descriptions or album notes on the platform itself.

When I started writing The Odhracle I did not intend to make playlists for each letter, but I realised that was the only way I could get people to test the music I was recommending, but it’s still why I primarily recommend albums and EPs.

---

Curation over Personalisation

Once upon a time, Spotify’s algorithm felt like a shortcut to global music discovery - I could not count the number of acts that I discovered through its “Autoplay” and “Song Radio” features. It wasn’t all algorithm-driven either; Spotify had an army of professional playlist curators, who would attend live performances of acts they were considering including in their playlist as if they were record label A&R, to “find out what songs people are singing along to in the real world”. The curated “Global Groove”, “Folk Fabrique” and “Poolside Disco” playlists were hugely helpful in discovering some of my favourite acts.

Unfortunately in December 2023, Spotify laid off 1,500 employees, a move which Spotify CEO Daniel Ek said disrupted Spotify’s day-to-day operations more than anticipated. Whilst Spotify did not publicly detail which parts of the company the lay offs affected, it has been speculated that playlist curation was impacted, and indeed in early 2024 Spotify included “AI” more visibly into its playlist creation through the introduction of its “AI DJ” and AI generated playlists features. This followed the introduction of “Smart Shuffle” in 2023, where Spotify’s algorithm inserts similar songs into your playlists, which is now the default play mode in some regions.

Around the time of the enhanced AI features’ introduction, I noticed the curated Spotify playlists I previously relied upon were now “personalised” to me - full of artists and songs chosen by the algorithm based on my listening history. The main problem being these are songs already in my library, undercutting the music discovery benefit of those playlists.

The personalisation is also oxymoronically generic - the same acts and songs now populate different personalised playlists, so where some playlists where once quite different, now I often can’t tell them apart, creating a passive listening experience.

This inability to escape algorithmic recommendations is problematic for a few reasons. Most people agree it’s bad that social media companies know what content to push onto your feeds to keep you scrolling, and in my mind it’s no different that streaming companies know what music won’t actively engage you enough to make you skip or change the playlist.

Also, despite what tech giants may tell you, I think the algorithms are fundamentally bad at identifying music that you might like, and better only at identifying what you already do or don’t like - this is not conducive to discovering new music.

For example I’ve been listening to a lot of drone and experimental folk for the last three years. Prior to hearing Sunn O))) on Mary Anne Hobbs’ Radio 6 show and the mainstream success of False Lankum (2023), I was largely unaware of these genres. I had never been algorithmed any drone music on any streaming service.

Similarly, I’ve recently gotten back into metal. If you knew me in my early teens, you would know this was my genre of choice for a time, but I left it behind around the same time as Spotify launched in the UK, and didn’t revisit it until last year. I’ve massively enjoyed going back to old favourites like System of a Down, and discovering acts that came along after I’d stopped listening, like Gojira. I haven’t once been pushed a metal act on any streaming service before now.

Both of these examples make sense - how could the algorithm know that I would enjoy these genres when they don’t correlate with anything in my listening history? Often music discovery is about taking risks and listening to something outside of your comfort zone, as you might end up loving it. Whereas the goal of streaming service algorithms is to keep you listening - Daniel Ek has been quoted as saying Spotify’s only enemy is silence - so why would the algorithm take a risk and recommend you something you may not like, that may make you switch off?

Engagement with music as an art form should be more than simply hitting the algorithm-targeted dopamine trigger in your brain as many times as possible - it should be explorative and thoughtful, involve taking risks to find new joys, not simply numbing the banality of life by engaging with whatever is most easily accessed. Music should evoke emotion, not be interchangeable with white noise.

---

Fake Artists and AI Music

Generic personalisation and naff AI features aren’t the only way in which your exposure to new music may be limited without you realising.

Journalist Liz Pelly, in anticipation of her forthcoming book Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist has published an excerpt titled The Ghosts in the Machine. In the excerpt, which I strongly recommend you read, Pelly sets out a few examples of Spotify practices you should be aware of:

• How “Spotify’s “Ambient Chill” playlist had largely been wiped of well-known artists like Brian Eno, Bibio, and Jon Hopkins, whose music was replaced by tracks from Epidemic Sound, a Swedish company that offers a subscription-based library of production music—the kind of stock material often used in the background of advertisements” - you may remember Epidemic Sound from Letter 5 in the context of Khruangbin copycat Arc De Soleil - companies like Epidemic Sound charge streaming platforms lower royalties than individual artists.

• Pelly’s investigation into artists who “had millions of streams on Spotify and pride of place on the company’s own mood-themed playlists, which were compiled by a team of in-house curators. And they often had Spotify’s verified-artist badge. But they were clearly fake. Their “labels” were frequently listed as stock-music companies like Epidemic, and their profiles included generic, possibly AI-generated imagery, often with no artist biographies or links to websites”.

• Pelly’s investigation revealed an internal program called “Perfect Fit Content (PFC)”, where it’s alleged Spotify encourages playlist curators to include those “completely made up” artists into its playlists (note the excerpt flags that Spotify denies encouraging its staffers to seed PFC music into playlists). Pelly asserts by 2023 “over 150 of these, including “Ambient Relaxation,” “Deep Focus,” … were nearly entirely made up of PFC.”

There are obviously lots of concerns about these kinds of practices in terms of lack of artistic merit and loss of authenticity, but what really horrifies me is that it’s laid the groundwork for AI-created music to fill niches currently filled by working musicians leading to lost opportunities and income - is listening to “Deep Focus” now the equivalent of buying a book written by ChatGPT? Pelly points out that Spotify has been “open about its willingness to allow AI music on the platform” and that “Daniel Ek noted that the boom in AI-generated content could be “great culturally” and allow Spotify to “grow engagement and revenue.””

I’m looking forward to reading Pelly’s book (UK release 13 March 2025), and if the excerpt is a a bit heavy reading for you, here’s a bitesize summary from The Honest Broker.

---

Payola 2.0?

I should briefly explain what “payola” is - historically it was a pay-for-play scheme where record labels would pay radio DJs to play specific songs, artificially inflating their popularity. Doing this without disclosing it to listeners is an illegal practice in the US; DJ Alan Freed was the first person convicted of it in the 1960’s, and in the mid-00s Warner Music Group, Sony BMG and Universal Music Group (UMG) were charged with it and settled out of court for millions of dollars.

In the streaming age where radio is much less influential on commercial success, and playlists and autoplay features are king, and the latter is where any modern iteration of payola would arise.

In an embarrassing death rattle in his beef with Kendrick, Drake accused UMG and Spotify of artificially inflating streams of Not Like Us through the use of bots. UMG and Spotify deny these allegations, with Spotify naming the only “Spotify for Artists’” tools purchased for the track - a visual ad promoting the song in France costing €500.

Drake spent much of 2024 embarrassing himself for the greater good (mostly our getting to laugh at him) and this lawsuit was no different, as despite it looking dead on arrival, it’s shone a light on streaming industry practices which are being referred in the music press as “Payola 2.0”, like Spotify’s “Discovery Mode”. It’s worth noting that unlike original payola, these practices are not illegal, and I’m less concerned about any legal grey areas and more about how they impact how we engage with music.

Using Spotify as an example (I’m confident other streaming platforms have equivalent practices), Discovery Mode allows record labels and artists to reduce the streaming royalties Spotify has to pay them, in exchange for increased algorithmic promotion on the platform. In her excerpt, Pelly says that like “the PFC program, tracks enrolled in Discovery Mode are unmarked [as being ‘promoted’] on Spotify; both schemes allow the service to push discount content to users without their knowledge.”

This lays bare the core problems with music recommendation algorithms; not only are they limited in scope by an incomplete picture of taste, and they are written with goals other than the furthering of musical discovery, but they are also influenced by commercial factors which are concealed from listeners.

---

The Virtues of Radio in the Streaming Era

The best way to find new music is not then in what an algorithm thinks you’ll enjoy, or through corporate committee-curated playlists, but in thoughtful and considered recommendations from someone who’s taste and integrity you can trust. I’ve found this in loads of places:

• newsletters (such as the one you’re reading… and others in the recommended list on The Odhracle homepage);

• podcasts (Nialler9, Song Exploder);

• review sites (Stereogum, Resident Advisor, BrooklynVegan & Pitchfork);

• online message boards (the Glastonbury-associated OOF discord and various subreddits like indieheads & TheOverload); and

• YouTube and Instagram (theneedledrop & Somewhere Soul),

but without question the best place to find new music is on non-commercial radio.

Around the time I got my first iPod I was fed up with BBC Radio 1 and its A, B and C playlists, and I largely stopped listening to radio entirely once I downloaded Spotify. It wasn’t until I started getting really into dance music in about 2016 that I came back to it again - first though Radio 1 Essential Mixes and Annie Mac’s Friday night show, later to Hessle Audio’s Rinse FM show and eventually into the world of NTS.

Independent radio can not only introduce you to new music that you’d not otherwise hear, but the DJs provide crucial context as to how and why the tracks they’re playing were created. This applies not only to new music, but also also you’re old favourites - I’ve had my mind blown about a song or artist I’ve listened to a thousand times by something a radio DJ has said before hitting play on it.

Radio is important, and finding DJs that you resonate with and trust to take you on a musical journey is something that’s so special, and it’s genuinely sad that because of streaming so many people have lost out on those experiences.

So I’ve listed some of my favourite radio shows below that I strongly recommend you check out. If none of these take your fancy, try sifting through the huge number of shows on online stations like NTS, Rinse FM, The Lot Radio, Soho Radio or your local independent online radio station.

Click the title of each show to be taken to the show’s webpage.

---

The NTS Breakfast Show w/ Flo - NTS

• When and where: Monday - Wednesday - 9am to 11am

• What kind of music: anything from funk, soul, indie-rock, alt-pop, jazz, ambient, disco & hip-hop.

Since taking over the 3 days of the NTS breakfast slow in the wake of decade long resident Charlie Bones’ departure, Flo Dill has made the slot her own. Her dulcet tones and eclectic taste are the perfect start to the early half of the week. Her exquisite tunes are only half of why the show is great - I love the regular segments like her agony-aunt role in matters of the heart, and hearing updates on her allotment.

---

Moxie - NTS

• When: Every Wednesday - 3pm to 5pm

• What kind of music: house, deep house, techno & UKG, and some R&B and soul.

My favourite radio DJ, I’ve not missed a show (or at least not listened to it on playback) since about 2018 - truly keeping it locked and keeping it Wednesdays. My favourite episodes are when she gets on her club and deep house tips, but her guests are always fantastic, platforming brilliant underground DJs and producers, I love that we share so many favourite acts outside of dance music like Cymande and SAULT.

---

Mary Anne Hobbs - BBC Radio 6

• When: Monday - Friday - 10.30am to 1pm

• What kind of music: Anything and everything.

Ok the announcement 2 days ago that Mary Anne Hobbs’ daytime Radio 6 show is ending in February and that she’s taking a 6 month sabbatical, is not ideal timing. She will be back with a new Radio 6 slot later in the year, and there are plenty of shows to catch up on on BBC Sounds. A legend of public radio, its arguable no modern DJ has been more effective at introducing mainstream audiences to non-mainstream music.

---

Bradley Zero Presents: Rhythm Section - NTS

• When: Every 2nd Wednesday - 1pm to 3pm

• What kind of music: House, funk, disco & deep house.

One of the smoothest voices in radio, Rhythm Section label-boss Bradley Zero never misses on his fortnightly show. Expect smooth rolling house music that’ll have you wiggling in your desk chair. Also love his occasional specials on regional music, like his Dominica Special from 2023.

---

The NTS Breakfast Show w/ Louise Chen - NTS

• When: Thursday & Friday - 9am to 11am

• What kind of music: House, disco deep house, funk, soul & occasional indie and shoegaze

I was gutted when Zakia announced she was leaving the late-week breakfast slot for the BBC, but was beyond delighted when one of my favourite DJs was announced to take her place. Louise absolutely nails the transition from week to weekend, and recently does so remarkably with her newborn baby often strapped to her. Delighted to have ChenTS back.

---

Gilles Peterson - BBC Radio 6 (Worldwide FM)

• When: Worldwide FM - Thursday 8am - 12pm & Radio 6 - Saturday - 3pm to 6pm

• What kind of music: jazz, funk, soul, hip-hop & electronica.

A legend of the UK airwaves, Gilles remains hugely influential through his radio stations, record labels and multiple festivals. His shows were an important entryway to jazz music for me.

---

Raji Rags - NTS

• When: Monthly - Friday - 1pm to 3pm.

• What kind of music: Hip-hop, techno, house, bass, electro & bass.

Love the variation that Raji’s show brings to NTS’ electronic lineup - his blending of hip-hop with club music hits the sweet spot for me. Also love the South Asian sounds he highlights on his show - the live set from Baalti last July a particular highlight.

---

Club Coco w/ Coco Maria - NTS

• When: Every 2nd Monday - 6pm - 8pm

• What kind of music: Funk, Latin jazz fusion, cumbia, samba & bossa nova.

Unrivalled at bringing hip shaking sounds from South America, Central America and the Caribbean to European airwaves. The Amsterdam-based crate digger brought her show from Worldwide FM to NTS, and stands out as the selector supreme Latin American records this side of the Atlantic.

---

Hessle Audio - Rinse FM

• When: Every 2nd Monday - 8pm to 10pm

• What kind of music: Techno, bass, jungle & garage.

The long running show hosted by the nice boys of dance music. The Hessle Audio label founders Ben UFO, Pangaea and Pearson Sound rotate hosting this slot. Each a legendary figure of the UK underground dance music scene in their own right, the radio show is one of the most consistently quality electronic radio slots around.

---

The Do!!You!!! Breakfast Show w/ Charlie Bones

• When: Monday - Friday - 9am - 12pm

• What kind of music: absolutely anything provided it was released prior to 2010.

Charlie Bones has been instrumental to my music taste, and his show (then on NTS), eclectic taste and larger than life personality was important to my mental health throughout various Covid lockdowns. You never know what to expect on Charlie’s shows, from playing entire live Nina Simone sets, to straight house and disco bangers for 3 hours, with every Wednesday entirely programmed by the chat room. I’ve kind of left Do You behind since the acrimonious departure of beloved early morning DJ OG a few months ago, but given the influence this show had on me, this list would feel incomplete without it.