Way back when Last.FM has its own radio service, I could throw a few random genres and/or artists at it, and it would recommend me pretty much exactly want I wanted every time. I gladly enabled scrobbling in my music players, and it tended to recommend me good stuff every time. Ever since its radio feature got killed, its database has been getting worse and worse. Just now, I tried to search for some things it used to be good at finding, and the artists section was not filled with artists at all, but rather a bunch of what appear to be random playlists with incomplete metadata.
Pandora had decent algorithms for recommending things, but it had such a small library that it would frequently repeat the same handful of albums for anything I searched for. This irks me, as I hate wearing out good music.
Spotify is currently where I keep my weeks-long playlists that I've built over the past couple decades. Even with such large playlists as input for their radio recommendations, Spotify doesn't do a very good job recommending new music either.
Whatever happened to the good databases and their algorithms? They definitely used to exist.
> Pandora had decent algorithms for recommending things, but it had such a small library that it would frequently repeat the same handful of albums for anything I searched for. This irks me, as I hate wearing out good music.
I've been a daily user of Pandora for something like 10 years. It's been getting steadily worse the whole time, and especially in the last two years.
I like to create my own station that is "seeded" by a few artists, and then allow the algorithm to do what it wants to play related music. This used to be great, until one day I noticed that it had become stuck playing the same 50 or so songs. This was after about three years of listening to that station on a weekly basis. As an experiment, I created a new station and seeded it again with similar artist. Again, it was fine for 2-3 years until it got stuck on a handful of songs. I did this again recently and it has become stuck within four months. There are even a few of my seed songs that it simply ignores and never plays.
Last.fm had the best categorizing features I ever saw. Songs would. It only be listed under “female artist” but would also have things like “extensive vamping,” that really got to the heart of what the song was about. They got closer to understanding why I liked songs than any other service. All of the other services feel rather dumb when it comes to preferences. It feels like they still rely on associating other people that liked artist x also liked artist y. As a music aficionado it’s infuriating.
Apple has tried to help by using human curated playlists but I frequently find myself thinking I have better taste than their “experts.”
I used Pandora a lot when I was a teenager, and experimented a bit. If you were listening the same time I was, the size of the library probably wasn't the primary issue.
Presumably their recommendation algorithm and the Music Genome Project tagging allowed them to, given a set of tags, find similarly tagged songs. This worked really well. It's how they used it and how they picked what track to play next that caused issues.
First, thumbs up. As far as I could tell, at the time Pandora would strongly prefer to play a track you'd thumbsed up on a station over anything else, as long as it was available to play, which mainly just required it to not have played to you in the last two hours. So if you used thumbs up the normal way, a well-used station would eventually turn into a loop of things you'd already heard.
Second, the way they used the algorithm - it seemed to me like they only or mostly used the station seeds as the input, there was no blending going on, and if thumbs ups impacted it, well, they had their own problems. That is, if you had a station with two seeds, it played some songs close to one seed, then it switched to the other seed and played some songs close to the other seed. Skipping would usually bump you to a different seed.
To get around all this and get a variety of new material, I created a station with a lot of seeds - 25 to 50 or more - and never thumbsed up anything on it.
> Whatever happened to the good databases and their algorithms? They definitely used to exist.
It was called a record store employee, and they no longer exist.
There's an AI search engine I'd like to see ...
Even the "market segmentation" of pop music still doesn't work for crap. Even something as basic as "Gee, I like 80s New Wave, how about recommending some artists born roughly in the 21st century who would fit?" seems to be totally beyond the pale of anything currently existing.
My tastes have always been niche enough that most record store employees would give me the "deer in the headlights" look or condescend to me when I asked for a particular artist or similar that I found on Last.fm.
If the database quality of current Last.fm were similar to its state back when it had radio, I would think that an AI trained on their data would be pretty good. With the current state of it... It would have to be crap. Heck, even if an AI model could be trained on the play counts of every song of every user on every streaming service, I'm not sure it could approach the curated relational algorithm that Last.fm had at its peak. Would definitely love to see an attempt, though.
I'm happy to not relive the days where CDs were and inflation adjusted $35 for about 10 songs. And there is no economic incentive or guarantee that any given retail employee would know anything about the inventory. Go to home Depot and canvas some of them about home repairs if you don't believe me.
> I'm happy to not relive the days where CDs were and inflation adjusted $35 for about 10 songs.
Limited edition vinyl stuff, for example, generally goes for right around that in order to support the artists. Most of that money is going directly to the artist, nowadays (as opposed to live in which it all goes to Ticketmaster).
If you're not willing to spend at least some money, well, then you're part of the problem why artists can't get paid for doing music and why so much of it kinda sucks.
> And there is no economic incentive or guarantee that any given retail employee would know anything about the inventory. Go to home Depot and canvas some of them about home repairs if you don't believe me.
Sure, if you went to Tower Records, you almost always had someone clueless. However, the point of going to those stores was to NOT go to the big retailer, it was to go to the local record stores that had people who worked there specifically because they were super passionate about music.
This was how you found out about that super obscure artist. It was also how you found out about the local bands that might be of interest to someone who liked that super obscure artist. etc.
There was an article in the Times recently about how Spotify is no longer solely using "how much do we predict the user will like this song" as a metric, but now is also considering "how much do we pay this artist per play" to optimize for cheaper higher-profit-for-spotify music.
Amazon music does an 'ok' job at recommendations for me, slightly better than spotify, but that's not a high bar. Jango does a better job, or did last time I used it, but it has so little available music that it somewhat self limits itself.
In the age of internet, engagement optimization and recommendation algorithms create a new way that we are affected by the behaviour of others.
That annoying dark pattern on a piece of software you use? Because there are people who fall to it, clicking on an ad or "engaging" more. That stupid show that keeps being recommended to you? Because a lot of people just sit on the couch, watching something on the list that does not need too much mental processing.
I have a peculiar taste in Music. I love many many different types of music, but once I find a really good piece, I'm not interested in things that are very similar to this one. Looks like if we describe musical work with high dimensional vectors, I like to find good vectors that are not too close to each other. But as the author said, Spotify keeps showing me music that's similar to what I've listened. That's exactly what I don't like (with the occasional exception of something being better than the one I've already found and replacing it).
I assume I belong to a peculiar minority. The recommendation algorithms work very well for predictable majorities.
Maybe someday we have an interesting "musical embeddings" model, and then people can implement personalized discovery algorithms using that?
I'm not sure how the Spotify recommendation algorithm works at all, but for some reason I imagined them doing fancier things than looking at my liked songs and finding similar ones. I would've thought they'd build a profile of you, and then find similar user profiles and show you songs those folks liked that you hadn't found yet.
That's gotta be how they do it, right? I'm probably wrong.
I don't work at Spotify anymore and I didn't work on the tech I'm describing, but I picked up a bit about what was going on while there.
First, there is/was no single algorithm, but the core ideas driving a lot of recommendations is:
1. Create user taste vectors
2. Match those vectors to other users or collections of tracks
3. Use that information and combinations of other things to find recommendations.
Each step of the process is constantly being experimented with. Different custom playlists might be using a different combination of tech doing those basic steps.
"Other things" including intentional commercial biases presumably?
No matter what I do in Spotify, under several different rounds of accounts, it always seems to gravitate towards the tastes of the general public, i.e. some form of mass-market pop.
Their recent "ai" assistant was a slight improvement because you can ask it for less popular music which is typically better for music discovery.
Collaborative filtering is similar but for huge recommender systems they’re not going to create a huge MxN matrix where M is users and N is items. I think what they’re referring to would be called a “two tower” model where you have a learned vector for the user, a learned vector for the song, and the cosine similarity is their affinity. It’s pretty performant because you can cache the song vectors.
I feel those are how Pandora and Last.fm (used to?) work respectively. Nowadays everything seems to just put a bunch of tags on a track and suggest you things with the same tags to the tracks you liked. Doesn't even need to match the same combination of tags, just some number of them. The problem is, you probably care about the small, specific tags, and the system cares about wide "popular" tags. If you like a couple niche genre covers of songs that happen to be featured in TV openings/OSTs, you are not getting more songs in that genre - you are getting a bunch of covers and OSTs.
I wish I had a music recommendation service built on Pandora's immense dataset of music tags that could build me a playlist that I could link back to whichever music service I happen to be using at the time. I could have it do things like require at least 3 tags in common between adjacent tracks such that it could jump around between 2 dozen genres but the transaction between any 2 given tracks isn't too jarring. It'd also be nice if I could tell it to make a playlist where every song shares one particular tag in common.
Maybe I'll build that. Sure would be nice to have.
The primary advantage of Pandora's algorithm is the human-labelled Music Genome database. I haven't seen any other company do music discovery as well as Pandora, and don't expect that to change any time soon.
I have no specific insider knowledge, but a decade+ ago they bought a company called the Echo Nest that was developing some of the best audio signal analysis algorithms around, I assume much of that influenced their recommender system.
Nowadays, they have a quite busy research department so I would imagine that recommendation is quite fancy indeed: https://research.atspotify.com
If you have any taste at all other than "maximally dissimilar to anything I have liked before," there should be a feature that predicts songs you would like.
If your taste is exactly "maximally dissimilar to anything I have liked before," that's actually pretty easy to calculate from the embeddings as well.
It didn't sound like op wanted maximally dissimilar from what they've liked before, but instead maximally dissimilar from what they've listened to recently.
> I'm not interested in things that are very similar to this one.
Are you sure this is an accurate description of your taste? Or do you mean "I'm not interested in things that are similar, but lower quality"?
I'm very much a weird-music enjoyer, and I often have the latter problem where "similar" songs actually just don't capture the same vibe as the truly engaging new song I just heard. But that's not because the algorithm is choosing music that's too similar; it's the opposite. It's trying to choose something similar but can't, so it just picks the next-best thing which I actually don't like.
Instead of relying on things like "Playlist made for you" I still like leveraging the algorithm to discover "new vectors" by actively going to a song that some aspect caught my attention and going to its radio (only tried on Spotify). Sometimes I'm surprised with new soundscapes.
Yess! Thank you for commenting. I am very interested in this topic. Please do share with me if you find any interesting ways to explore new music for your taste
Same as GP. It takes times. Unless it’s party mode, I only listen to albums. To find new music, it takes time mostly. I decide to listen to a new genre and I seek a playlist or a compilation curated by someone. If I find someone I like, I check their albums. I also checkout the recommendations on bandcamp (people vote with their wallet there). Then there are forums and polls, and I may decide to try something out of the blue.
The more you curate, the more you define your own taste. It’s then easier to describe what you like in a music and triage.
This probably won't work for most people, but as I mostly like metal, I simply check all metal releases that seem slightly to my tastes according to the subgenre. Then when this sampling seems at least somewhat positive, I play the full album after release. That usually ends in sampling songs from 30 albums, listening to ten albums, and buying 1-2 every Friday. Never had an algorithmic recommendation system that worked for me.
local record store. talk to the people who work there, tell them the 5 albums you’ve been hooked on lately.
at the end of the day no matter how many times we beat our heads into the same wall, we’re not even close to an accurate discovery model, music nerds are far better at recommendations than any discovery models. far better.
don’t let their insufferability discourage you. you will be too once you start diving into and going on rants about music which is outside of mainstream fluff. it’s like this with any $subject involving wonks. we’re insufferable to anyone who isn’t into our particular genre of technology. food wonks are insufferable, car geeks are insufferable, gamers are insufferable. that’s ok, if you’re looking for someone who is a geek in a topic, you’re likely to become one too :p just be normal around $subject non-wonks and you’ll be fine.
but yeah, music nerds working in a good record store really do know their stuff.
other places:
- music nerd streams on twitch
- music reviewer youtube channels
- college radio stations (most have an online presence) 770 radiok out of minneapolis is incredible
- kexp out of seattle is absolutely amazing (they’re heavily online as well.)
- just about every mid+ sized city has some amazing radio, usually found in the low FM areas.
at the end of the day though, it’s other people. there are far too many variables for every individual which drives why they may or may not like a song at any given moment. other humans are still absolutely unmatched when it comes to navigating this.
> local record store. talk to the people who work there, tell them the 5 albums you’ve been hooked on lately.
Unless your interests are niche. A fun game I used to play as a teen was going with my parents to the record store and seeing if they had any music I listened to online while my parents shopped. Never found a single CD (but they couldn't be that niche, this story is about bands I found out about from my friends at school!). Employees tried to be helpful, but there's only so much they can do when someone comes in and asks for a list of bands they've never heard of.
I agree strongly. Listen to the radio stream for any reasonably niche artist (especially if they have had a single "hit") and you will be treated to 3 or 4 different mixes of the same song within the first hour. For some purposes Spotify is just unlistenable.
This is one of my main complaints. Back when Google Play Music was still a thing, I uploaded my library of ~30k tracks that had been ripped from the massive binder of CDs I'd accumulated over the years, plus stuff from friends, stuff I got off torrents to replace albums I'd already bought on tape, etc.
They have a huge sample set of stuff I like just from that (it's still accessibly via YouTube Music, but it sucks). Still, any time I fire up YouTube Music and play a song, the next 2 or 3 can be solidly appropriate and often stuff I wouldn't have thought to play - in a good way.
Then there's a sharp decline where it just starts playing the same thing it played last time or the most common song off the most popular album from the most well known artist adjacent to what played previously.
The whole reason I use the "algorithmic feed" is because it should be optimized for my tastes and what will keep me listening. I still "thumbs-up/down" stuff but it never seems to work. You would think with their insane library and huge dataset of my prefs and listening data they'd be able to generate something great.
Instead I just go back to the human programmed Shoutcast stations I have bookmarked after a few songs.
The same for me. Nowadays i make my own songs using Udio and i upload 'em to YT. It is unlikely i will ever listen to suggested songs ever again, by any service.
So I've been using Tidal for 5 years now, and feel they run circles around Spotify in terms of curation. Their algorithms and curated tracks are better, and they steer away from the social/gamification features and lean in to artist-centric features. For example, they've had a "credits" feature since day one - you can look up the producer, guitarist, oboeist, etc of any song, and see what other work they've done. In terms of discovering new music, there is absolutely no comparison to being able to look up the actual human behind the track. I've discovered hundreds of bands this way.
In my experience, and the reason I left Tidal several years ago, is that they lean heavily into modern hip-hop, compromising relevancy for the sake of promoting "friends of the company" (Jay-Z, Beyonce, etc).
Spotify only lets you read a very reduced version of the credits of a song, and you can't click on the artist to see all of their work. On Tidal, the credits are much more extensive, and you can go deep down the rabbit hole of a producers work in one click.
More importantly though, Spotify doesn't directly link artists to bands they're in. For example, on Spotify, Billie Joe Armstrong is credited with his work on Green Day, but if you go to the Billie Joe Armstrong page, he isn't. On Tidal, each artist can also be broken down by their role on a project. For example, you can see just the songs that Paul McCartney wrote (turns out he wrote a track for Drake called Champagne Poetry in 2021... who knew). On Paul McCartney's page on Spotify he isn't even attributed to The Beatles
> For example, you can see just the songs that Paul McCartney wrote (turns out he wrote a track for Drake called Champagne Poetry in 2021... who knew).
Unexpected writing credits like this usually indicate sampling. In this case [0], Drake sampled a song [1] that sampled another song [2] that was a cover of a Beatles song [3].
I used Tidal twice, for a year or so each time, and I found that their recommendations were the best by far. But their app was so insanely buggy. On multiple different phones it would have to be force-closed and restarted about hourly. Usually it froze when starting a new song. The whole screen would be frozen and unresponsive. Now I use Apple music and the recommendations are garbage, and it also freezes frequently, but specifically only when I wake the app up from sleep.
Importantly for music discovery, Tidal make it relatively easy to browse by record label so if you're not listening to major label pop stuff you can pretty easily find a bunch of artists in a similar niche. If that is possible on Spotify they've made it hard enough to access that it might as well not have been there.
Last.fm was good, but inevitably went downhill when it was bought out. As they had to commercialise more and more from the early days as a uni project that became audioscrobbler, with a tiny userbase, they followed the standard Doctorow model of capitalist decline. They had a sweet spot a few years in when they had plenty of data flowing and they hadn't yet messed up their APIs. Right now I use it to log my listening but I'm waiting for the email that says I'll need to move my data elsewhere.
I would naively assume that both Spotify and Tidal pull that information from the same sources. Hopefully there is a web service. Not to take anything away from the first person who decided to, you know, actually tell us this information. At this point it should be table stakes for a music streaming service-indeed, it should be a legal requirement if you ask me.
Thanks, that's really surprising. Credits get into ownership and payment, and I assumed some group of authorities would have that on lock, and would have pooled all their data since it's sort of in their interest to do so. Silly me I guess!
Spotify just pays the owner of the song (typically a record label) so doesn't need to be concerned with who created it. But yeah, even then, it's surprising the Spotify credits are so poor. Everyone forgets about Ringo!
I think I speak for most people when I say I want my music "discovery" playlist to be something like "mostly stuff that sounds like stuff I've indicated that I like" with a small amount of "not sure I'll like it, but suprise me". It sounds like OP is looking for the ability to turn up the "surprise me" factor at the expense of maybe having to skip a few more songs that just aren't clicking with them. So maybe something like a "temperature" knob is in order?
"They are not suggesting new, very interesting melodies. They are finding you the tweaked versions of the songs you already like and, even on your first listen you can predict the melody that’s to come."
I really don't think that's the main method of Apple Music or Spotify to create a list of suggestions. From what I know, (beside of dark marketing-patterns) the suggestions are created by checking what other songs people like/listen to who ALSO like/listen to this current song (or other songs you played), and the common neighbors of those songs in other playlists.
(If you play music for your toddler, your future suggestions will include children's music not because it sounds similar but because "a critical mass of other people who listened to Baby Shark on repeat also listened to: Old Town Road")
It is weird and it’s ironic that they call that “discovery”, as it feels more like variations of what I'm already listening to.
This indicates that the persona that this platform created for you is quite homogenous and probably matches closely with many other personas on the platform, so many people who listen to the same music as you do apparently listen to _nothing else_ than this kind of music...
(not trying to defend those suggestion algorithms, just analyzing the comment)
> This indicates that the persona that this platform created for you is quite homogenous and probably matches closely with many other personas on the platform
To add to this analysis, I think there may also be a feedback component to this problem that exacerbates the issue, since most users are passively using the suggestion algorithm.
In other words, if the suggestion algorithm tends to create a homogenized persona of the user's taste, say, because they don't bother to actively correct it, then this persona is embedded into a cluster of people with similar personas. And because the persona is now closer to said cluster, the suggestions will become even more homogenized. Moreover, since the cluster is mostly composed of passive users, the cluster itself will tend shrink (eg in variance) and to get more homogeneous.
I suspect that most algorithms do not do enough to prevent this global trapping effect, and so even if they have some method to sample "something new" for the user this becomes less and less efficient as more users rely on the algorithm for their suggestions.
I actually do find this observation to be quite accurate for many of my own 'suggestions.' I'm regularly recommended 'new' and 'old' music that was clearly matched to my 'tastes' only by melody or, more noticeably, sample. It very much seems like if a song fits into a genre I listen to frequently or have been listening to lately, and it samples another song I've listened to before—cheap recommendation. And the greater the frequency of individual plays (i.e. the more times I've replayed any one song), the more likely that derivatives will be recommended to me.
It's easy to see how this would've been baked into a human-made algorithm when you consider waveforms. Speaking only to Spotify's algorithm here. And it doesn't really bother me for obvious reasons. But it is creating something of a musical echo chamber for me.
Even though the idea of recommendations is anything but new, literally nothing and nowhere works as expected. The only thing that comes close is based on the concept of neighbours, as implemented at Last.fm or RateYourMusic.
I don't understand why is it so hard to offer something along these lines:
1. Define dominant user preferences by clustering and segmenting the field of listened genres.
2. Build a list of relevant "neighbours":
2.1 Manually added users/friends
2.2 For each of the dominant genre preferences, find users with a high level of artist intersection within that genre and add them
3. Now, for a "find similar" query:
3.1 Define a reasonable time window
3.2 For each neighbour, find points in time when they listened to the queried track/artist
3.3 Build a list of tracks/artist from the defined window around the points found
3.4 Filter tracks/artists that are too "distant" on the general genre/tag map, or lie outside of the user's dominant preferences (with a degree of boundary feathering, perhaps)
3.5 Filter if similar to negative part of the query
3.6 If novelty is required: filter artists/tracks according to the degree of their presence in the user's history
I think the financial incentives to promote specific artists (or songs with ads in them "GUCCI!") become the focus pretty quickly.
The cost to play a song is expensive, so if you can actually profit by putting a new artist instead of paying, why wouldn't you?
Sure your customer gets 3 minutes of potential garbage, but they don't realize that they generated revenue for the company just by sitting through that song.
If you give your customers a great experience, they are going to listen to more music, which is bad for the bottom line.
There doesnt seem to be any competition due to IP laws, so there is no incentive to be good.
1. The curse of dimensionality when computong distance functions and
2. The cost function of a bad song. People get mad if they get too many things they don't like. Notice Pandora immediately sends you back to the band that seeded a station upon any thunbs down.
The two issues I've had with every discovery algorithm:
"We have [favorite band] at home" - it picks things you like from your favorite band - instruments, tempo, etc then finds bad knockoffs that are superficially similar but painful to listen to.
The "Iron and Wine" problem - some bands are so generic that they tick every single similar box and flood your recommendations. For years, it didn't matter what band/genre I tried to find recommendations from, I got Iron and Wine.
I think a more fundamental problem is that people like music for very different reasons. Even the same person may define "similar" very differently at different points in time.
If I want music "like" "Groove is in the Heart", is it because:
* I want mid-tempo house-like dance music
* I want major key songs with female singing
* I want songs with rap interludes
* I want 90s music
* I want fun party music
* I want music that reminds of that awesome trip I took with my friends a few years ago where we played a bunch of songs over and over
There is no right answer to this question. But, outside of just looking for playlists, no music app I've seen gives you a way to specify in what way recommended music should similar to the current song.
I see this effect most acutely when I listen to something that happens to be popular. For many people "heard it a lot when doing this fun social thing" is one of the main reasons they like a particular song. This was true for me too when I was younger. But for me today, I'm mostly oblivious to popularity. I just like stuff that sounds a certain way.
Whenever I stumble onto a song that has a particular sound I like that happens to be well-known, the recommendation algorithm just starts throwing other popular stuff at me that sounds totally different.
Ironically, this is something that I think Pandora solved quite well with their recommendation engine. By virtue of creating a station around a particular vibe, even if 5 playlists all started with the same seed song, weighting other songs up and down on each station would curate a different listening experience, by virtue of finding how those are similar. Where Pandora was limited (at least, the last time I used the service) was the pre-seeding process is a bit arduous and opaque. I'm not sure how you make that easy to interact with, as going a layer beneath to the "why" a song was recommended and allowing folks to influence the graph at that layer sounds like a daunting UX challenge.
> it picks things you like from your favorite band - instruments, tempo, etc then finds bad knockoffs that are superficially similar but painful to listen to.
This is pretty bad if you have strong feelings about how much screaming a metal song should have. There are songs that fit exactly what I like except for that variable and Spotify does not get that I keep skipping them for a reason. It's rarely a "bad knockoff", but it definitely hits "painful to listen to".
It's really strange to me since it successfully creates playlists around different types of music that are sort of similar but shouldn't cluster together.
One of the things I find frustrating about music suggestions is that the app/algorithm doesn't care why you like a certain band. A long time ago I asked Pandora to make David Bowie station and it rolled me a generic classic rock station -- Zeppelin and the Stones. I was hoping for old school glam, maybe T-Rex and Eno. There's no way to communicate that desire to our music players. To say, "please don't think me a basic AF music listener."
I have noticed an interesting phenomenon around TOOL. If you start a playlist on Apple Music from TOOL it will start playing everything from Metallica to Nirvana. A lot of people like TOOL for a million different reasons and Apple doesn't know any different except for the overlaps in taste. If you play a Mike Patton band, such as Mr. Bungle though -- you will get some TOOL in your playlist -- because both bands are esoteric and often challenging.
I'm looking forward to the day (or wishing maybe) when my app considers these factors. For me the issue isn't discovery, but rather I want my robot DJ to vibe more closely with me.
It seems some of these services (e.g. Spotify) don't really do musical similarity, but instead emphasize indirect "other fans also like" similarity.
That tends to disregard many reasons you like a particular track, and does especially badly when the liked-track isn't part of a uniform style for an album or artist.
I recognize it's a heck of a lot easier to implement, but it's still a disappointment.
They definitely do both, in the public recommendations API you can see vestiges of the old EchoNest acoustic properties along with some new ones they’ve come up with. It’s fun to play around with.
He’s quite biased towards not using ML or acoustic characteristics for recommendations. But even if you disagree it is interesting to hear about how things were working under the curtain (for daylist in this case).
At its peak, Last.fm was pretty good at this. It had enough data on "people that like these songs also tend to like these songs" that it would generally know what to recommend. I miss those days...
Personally all I want out of a subscription based music service is excellent quality, constantly updated/created, thematically consistent, human curated playlists. I don’t really care if it’s super popular or fringe stuff, but I do want it to be as “good” as the other stuff on the playlist, and I want a human who also cares about the music to be making that decision. Sometimes the playlists in Apple Music scratch this itch, but it would be amazing if they were constantly updated.
Deezer has human curated playlists updated frequently, which has been my main way of discovering new artists these days. For example, the prog metal playlist as been updated 5 days ago. [0] There are lots of those playlists..
I've got a Tidal subscription, and I've played with creating a few of my own playlists for the explicit purpose of sharing with others (as opposed to personal use). They're nothing special, but I at least try to put in some research while constructing them to bound them at a narrow time-frame/genre for accurate historical purposes.
I have no idea if anyone is listening to them, though, because there doesn't seem to be any feedback system for the community playlists. That would be a useful addition, IMO. If TIDAL doesn't want to pay dedicated staff to curate playlists, they could at least make some way for the member-created playlists to get featured or gain reputation.
No music discovery algorithm has satisfied me. All data-driven approaches make predictions based on historical data. Personally I enjoy being exposed to entirely new genres and sounds I've never heard before, instead of variations on genres I've listened to a lot.
My solution: listening to NTS, an eclectic online radio station, where diverse artists create playlists.
I agree that NTS radio is one of the best ways to be exposed to interesting new and old music, obscure stuff, brilliant mixes, etc.
The NTS app is great: for Web, Android, iOS - it's always being steadily improved. A very nice feature to aid discovery/curation is that every track in a tracklist has a 'copy song and artist info' so you can easily search for tracks on your streaming platform. Not sure if this is a subscriber only feature.
I also use the 'identify song' feature in the Google search app on my phone, similar to Shazam.
If the algorithms aren't doing it for you then do yourself a favor and head to https://nts.live
If entirely new things is what you're looking for, you're not really looking for a recommendation algorithm [1]. What these algorithms try to achieve is finding unknown songs that are in the same genre to what people already like.
[1] Technically "random song not in listen history" would work out, if you'd really like to call that a recommendation algorithm.
But I also don't want a totally random song eighter. I want something that vibes with me but not directly recommended through my listening history, because then they are extremely similar and feels like they're feeding me the same melodies over and over.
Thats why I tried to give the "vibes" in a different format; image, rather than my listening history.
Radio Paradise is very much a rock station at heart, so necessarily for everyone's liking. If you're into classic rock mixed with contemporary rock, mixed with a bit of everything else, it's worth a shot.
In my experience literally anything beats that one friend who is a DJ. My friends - professional DJs in Berlin - haven't even heard of Can, to my absolute shock.
The “and still obsessively digs crates” was important I think. DJ’ are not even close to fungible (hell, I’ve even met a couple who don’t like music much)
All music recommendation engines at this time still aspire to be mediocre, they aren’t even playing the S.A.,e game as a human who is good at it.
Unfortunately, such humans are unevenly distributed.
My point is more about DJ specialization I guess. The vast majority of DJs I know personally still dig crates all the time - but they are techno/house/D&B/etc DJs, they know close to nothing about music outside of these genres. This goes so deep that some German techno DJs haven't even heard of Krautrock, the German scene that in many ways was a precursor of electronic music.
German beat-oriented DJs haven't heard of one of the most influential German bands that had a critical impact on beat-oriented music? Just an odd thing for people who dedicated their life to music.
Pandora, Pandora, Pandora. No other service for me worked to reveal new songs and artists I did not know about that nailed my taste. But Pandora relies on manual tagging by music experts. Fantastic, but probably not very scalable.
Grooveshark. Pandora was miss-or-ding-the-side-of for me. Grooveshark always seemed to grab me songs that I NEVER would have discovered on my own, but that resonated with me somehow. It was also a lot easier to find particular versions of songs, or indie music (particularly before Soundcloud became popular). The benefits of not having to kowtow to music industry IP BS (until it was destroyed by it).
I was about to say that I've been pretty ok with Pandora so far. My wife thinks I "use it weird" as I mainly listen to the Shuffle station and have mostly various artists in my collection. Whereas she has styles and categories. And that's mostly because I don't want to listen to "Mid-90's Grunge", I want Soundgarden, Pearl Jam, etc.
However, even in the various artist stations, they do play artists "similar to" as well. Which is how I started listening to stuff like Murder By Death, Eilen Jewell, Bakar, Black Pumas, and others.
This is great to hear, though I'm curious why photos on your phone / pinterest would be relevant to a recommendation system? Surely the biggest signal would be what Spotify already uses: the features of various relevant factors (your previous listening sessions, your current session, what other similar sessions look like, etc.), that said, their recommendation system is surprisingly terrible given how much easier music recommendations must be versus video, yet YouTube seems to have had this nailed for 15+ years whereas Spotify's "Discover Weekly" is so bad.
> I come up with an idea of generating playlists from images. Images that you shot on your phone yourself, or that you found on Pinterest, or a painting that you really like and feel inspired by.
This is genuinely interesting! Do you send the images to an LLM with a prompt like "generate a list of songs that would go well with this"?
I think spotify's playlists are quite decent. I think they used to be a lot better than they are now, but I suspect that most recommendation systems decay over time. I suspect they don't handle the recommendation feedback very well, so they can start off introducing people to new things, but then become a bit more static and just reinforce the same habits over and over.
I think YouTube's recommendations used to be excellent, especially for music, but I've personally found it to be terrible recently. It no longer recommends anything new to me, and I suspect that it's way over-tuned. If I see a video that looks mildly interesting I'm a bit hesitant to watch it, because I don't want YT to decide that it should become 50% of my feed for the next week. Which from the recommendation system's perspective is just weakening the signal I'm feeding to it even further.
This is an interesting point and I wonder if part of the issue is that a mature, extremely popular algorithm trends towards the lowest common denominator. I don't mean this as a judgement of taste, it just seems to me that people engage with art in different ways. Maybe the spotify algorithm is perfectly tuned to the majority of people who just want to be able to find more songs that fit the kind of sound they like or find something to throw on in the background. But for a significant minority of others like myself and OP, it's just not tuned to what we actually want from new music.
I also feel the same as you regarding the youtube algorithm. I actually get better recommendations sometimes by just logging out since it will try showing me new stuff.
One thing I'm not sure about is whether it's actually the algorithm's fault or if my expectations have become unrealistic and made me lazy. I used to read magazines and blogs to find new music. There are still tons of people writing about their favorite music, labels that act as curators, etc. I just don't seek them out and instead expect to be spoonfed by the algorithms. Even if this is true though, I suspect many of these algorithms could do a better job.
Also RIP Netflix's old recommendation system. I guess it wouldn't make sense when they can't license every movie like they used to, but I remember it being great. Although maybe it was just pretty good and I was younger and less familiar with the back catalog of good films.
That's the other thing I wonder - am I just getting older and less excited about new things? There used to be a real vitality to finding something new and exciting. Now it kind of feels hard for anything to feel that fresh anymore, it all seems like variations on the same core ideas. I do still find new stuff that I like, but it doesn't have the same thrill. Maybe I'll always be chasing that dragon of youth haha.
Thank you! I should check youtube's algorithm too.
Yes, I have a prompt like that the current prompt is this:
'You match the vibes of the pictures with the right songs and turn them into a 3 song playlist with a playlist name. The music genre of the playlist should be consistent for each song. Be creative with music selections, explore different music, be consistent in terms of the genre of the 3 songs. The playlists should be provided in an object array format, like this: \'[{playlistName: "string", songs: [{songName: "string", artist: "string"}, {songName: "string", artist: "string"}, {songName: "string", artist: "string"}]}]\'. Do not add any other text information and only give outputs in the provided format. Your playlists must match the visual vibes and maintain the specified format without any additional information.',
Pretty basic, as i said before this was only ment for me and to explore this idea, but i loved it so I wanted to share :)
YouTube’s algorithm isn’t very good for users because it doesn’t really separate mildly interesting videos that you finish from awesome content you loved.
YouTube of course doesn’t care because they don’t make more money when you see something awesome.
YouTube didn't help things any when they neutered the dislike button. It's still there but functionally useless. Yes, we all know why they did it to save the feelings of a few political staffers at campaigns aligned with the values of workers at Google but it's been an absolute godsend for scammers and terrible for signaling interests to the recommendation engine.
That makes things worse IMO. My favorite videos tend to be one-offs not channels producing regular content. Unsubscribing from everything definitely improved my feed.
It's not the only way to discover new music, but in the age of 99.9% online presence, I feel simply going out to listen to music is way overlooked.
Most cities and metro areas with more than a few hundred thousand people have jazz, punk, indie, hip hop, country, choral, and classical scenes. Certainly true of any ville with a university.
Check out a local weekly, listen to college radio, look at the online calendars of local venues and clubs, take a risk, check something out you've never heard of before. You may be surprised. There are musicians and scenes which fly under the radar of widespread Spotify and Youtube popularity which nevertheless deliver great performances and often themselves lead to other new, interesting discoveries.
A side effect is you may also end up talking to someone at these gathering places and making new acquaintances: again, another great way to discover music and other things.
There is so much that has been built online whose subtle or sometimes overt goal seems to be to eliminate actual human contact. I suppose that is attractive for some, but I feel the opposite is what a lot of people yearn for, and it can be achieved with just a little bit of investment.
I've found YouTube Music's recommendations very good. I somewhat routinely do 14 hour drives and I always end up hearing three or four new songs I love. As I've listened to various songs it's done a great job of figuring out what within the genre I'm listening to I enjoy and don't enjoy. The idea of knowing the next melody, as the author says, doesn't really bother me if I like the track, though my taste isn't very diverse.
Many of you asked how it looks, how it works. I've added a video under the blog post, showing how it works.
Many of you were also interested in knowing what's going on behind the scenes. The setup is pretty basic. I've built a NextJS app, for the LLM model I am using open AI gpt-4-turbo and sending the images there directly without any database for images. I did a little prompting to get the same output everytime and when I get the output I make search on the Spotify API, find the songs and create the playlist with them on your own authenticated spotify account.
Likewise I also don't have a database for the emails eighter. I am using spotifys authentication.
As I said before this was just for me at first but it is very exticing to see many people interested by this idea.
If you don't trust openAI models and don't wanna send any pic data to them please don't write your email
> They are not suggesting new, very interesting melodies. They are finding you the tweaked versions of the songs you already like and, even on your first listen you can predict the melody that’s to come
This seems like the complaint of somebody who hasn't been using spotify very long. After a decade plus, I feel like my algorithm is a rich compost pile of all of my previous phases of music. Spotify is excellent at letting me broaden my horizons or jump down a rabbit hole from a random starting point, like a song I hear in a public space or commercial or something sent by a friend. Maybe the OP should keep their ears open to more sources of randomness from the outside world?
I feel the opposite: my Spotify recs (after at least 8 years with an account) tend to get stuck on whatever I've been listening to recently. I've had to consistently go afield to find any new (to me) music. Even their "new releases for you" falls short of recommending me releases from artists I follow. How much less capable could it be?
Release Radar is consistently the worst feature of Spotify. It misses entire new albums from artists I listen to regularly, and seems to have a quota of songs to fill so after the first two or three it's no longer aligned with my interests. I can forgive it not being coherent since it's supposed to include multiple genres together, but I can't forgive it going way off from what I like just to hit 30 songs.
Huh, I don't even have that section on my Spotify. I have a "New music you need to hear this week" at the very bottom (none of it is anything I need to hear this week), but it's just generic "new music in X genre" playlists.
This is my experience as well... I have a very broad music taste but with some main themes. I find Spotify's algorithm (11 years of Premium) to regularly surface things I'll like, whether new music from artists I already know, music correlating strongly with known tastes, or every once in a while something that seems out of distribution but I like it anyway!
It probably helps that the strongest areas of my taste are relatively small or niche genres, like Scottish trad and Celtic (folk) rock. In those niches, similar-but-different is often distinctively different in actual experience. Sure, there's covers of the same song from time to time, but I actually do like enough of those not to be bothered, if they bring something new.
Interesting. For example how would one find a "new" Tool or Deftones? Current algorithms probably don't "pick up" not-yet-so-popular things. For example Shelton San (I found out about them via word-of-mouth), although I'm frequent user of Spotify.
This means that classical promotion channels are still necessary, as otherwise things get lost in noise.
I noticed that Spotify surfaces similar artists who are also of a similar popularity. So it's not like it doesn't understand that particular style, it just has to somehow pick a couple of dozen artists to show in that very coveted spot.
So what worked for me in the past is finding less popular artists and then checking their similar artists.
I'd be quite happy if Spotify just went away. The world of music would be much better off.
My solution (doesn't work for everyone): I have a large library on the microSD card on my phone, and set the music player to Shuffle. Quite often a song comes on and I think, "Wow, I own THAT??"
OK, I'll admit that doesn't play any new music. However, no bills for bandwidth!
> I'd be quite happy if Spotify just went away. The world of music would be much better off.
Why? It's almost exactly the same experience you get from all the competing services, and that experience is fantastic: any album, any song, instantly ready to play.
> I have a large library on the microSD card on my phone, and set the music player to Shuffle.
This is a fine situation to be in but from the perspective of your fellow music consumers it is highly undesirable. Music libraries involve a lot of time and money. For the cost of one album per month you can subscribe to an unlimited service.
> no bills for bandwidth!
This definitely sounds unique to your geographical situation.
I admit that streaming services are bad for artists who want to make money from album sales, but asking consumers to use something else would be like asking people to ride horses to work in order to keep farriers in business. The 1990s are long gone; being a famous musician isn't guaranteed to make you rich anymore. Meanwhile the not-famous musicians who make up 99.999% of the music population can happily continue not making any money off of album sales the same way they've always done.
The most interesting thing in this, to me, is just how differently people perceive and enjoy music. I find that the algorithms work pretty well for me. But novelty in melody or rhythm are not something I care about (I literally can not remember an instance of even thinking about how a melody in a new song might go, let alone predicting it). The kind of “newness” that the algorithms provide work exactly right for how I enjoy music. But it makes a ton of sense that if you are more driven by finding new melodies, they would not work well for you.
The real big lie is the idea that the recommendation algos were ever actually for the user.
The premise that many folks miss here is the idea that Spotify is, at best, thinly interested in recommending music that is good for YOUR interests. Spotify is the music business, and specifically the pop music business, has long discovered that's it's much more economically expedient to force feed musical taste onto the public than it is to chase the whims of organic hit-making. Payola is as old as recorded music. Spotify recommends what Spotify wants it's users to listen to. They have all kinds of side deals and marketing deals with labels, they have cheaper costs/royalties on some tracks than others. Popular tracks cached in their CDNs are probably cheaper to recommend than long tail ones etc. They have strategic priorities like gaining on apple for podcasts, and therefore injecting allsorts of podcast recos in the UI wether you asked for that or not.
Music discovery has never worked for me, for the simple reason it's the lyrics, not the music. I listen to anyone speaking truth, (the truth I believe, of course), and that gives me a wide disjoint range of music, but they are all singing about political social truths. Marvin Gaye, Public Enemy, Rage Against The Machine, Beyonce, The Stranglers, The Jam, Psychic TV... It's the lyrics, and now today, we finally have the ability to have music discovery with the lyrics, with the intellectual content of the music and not just the dressing.
I only pay attention to melodies and rarely can even discern the lyrics unless I have them written out as I listen. Even relatively clean vocals like Johnny Cash just wash over me without being understood. The words for me become just another instrument that can play notes. If I can always predict the next series it is boring, if I never can, it is too challenging. In the middle I get dopamine whether I predict the notes or not. The right series of notes can make me feel such varied emotions with no words necessary.
Music discovery used to work for me with pandora, nothing else has. I have no idea if they had better algorithms, or just a better catalog. It doesn't seem to work as well as it once did either.
Tl;dr I don't think it is simply your preference for lyrical content over melodic content that causes the algorithms to fail. They are just bad.
That's a very good point, I also love discovering with lyrics. What I tried to do is to find some connections between the image and music, maybe some connection we as humans cannot see right away.
You might like folk punk - check out Pat the Bunny.
I feel like there are two kinds of singers - people who are good at singing, and people who have something to sing about. You and I, I think, prefer the latter.
Even though you'r being mean I feel the need to defend you, only because I agree that this was one of the most unrelatable things I've ever read on the site (i.e. GP's preference for music with a political or social theme in the lyrics rather than a specific genre, theme, sound, style, or mood).
If you're reacting to the inclusion of Beyonce, I suggest checking out her recent Country music album, which is a political re-envisioning of Country music and to many a direct attack on the comfort seat of White racism.
The setup is pretty basic. I've built a NextJS app, for the LLM model I am using open AI gpt-4-turbo and sending the images there directly without any database for images. I did a little prompting to get the same output everytime and when I get the output I make search on the Spotify API, find the songs and create the playlist with them on your own authenticated spotify account.
Likewise I also don't have a database for the emails eighter. I am using spotifys authentication
I don't have evidence for this, but my working theory is that there is too much money to be made in promoting certain artists / songs for music recommender systems to remain objective. If supermarkets auction off premium shelf space to the highest bidder, what's stopping Spotify, Tidal, Apple etc. from doing the same? Want to create a new pop star? Hand over the cash and we'll put it in millions of people's "Discover" playlist.
These popular playlists are often already owned by the music corps. And I am sure there is a price tag for those that Spotify manages.
Next to that there is also the shadow fact that there are millions
of dollars and more in fake plays, subscribers , ... SMM panels are getting big because the music industry doesn't allow for anything less than instant fame.
I really wish Spotify or Apple offered the ability for the listener to simply listen all of the songs released on their platform on a day or a week, good or bad, and directly pay the artists for the songs listeners enjoy.
Spotify's "New Releases" for example, tracks only music that major labels promote, or that fit a predetermined genre, or are similar to the songs and artists that you already listen to.
There are smaller services yes, which allow for independent promotion and distribution (Last.fm, RateYourMusic) but these have fairly obvious flaws in how the listener can approach new music (RYM pushes ratings first and foremost, and both last.fm and rym push trending artists to their users).
Instead, because the value of music is zero (really, negative since the number of listens, streams, album purchases, etc can fail to recuperate the cost to make it ), the act of distributing music presents economic risk unless the release itself can be controlled by the investors through advertisement or paid promotion.
Music discovery -- or discovery of movies, TV shows, love interests, or anything else that caters to human preferences -- is an enormously challenging problem that those whose business it supposedly is to solve have largely given up on. Part of the problem is exemplified in this thread: there are endless different ideas of what the ideal $ITEM discovery algorithm should do. Even if the Spotifys of the world were to deploy some modern AI-based tools to fix the problem, the level of tweaking each listener/watcher would need would be so extensive as to belie the entire effort. So they don't, for at least 2 reasons:
First, it's hard, as we all know.
Second, it turns out that the "best $ITEM discovery algorithm" accolade doesn't pull in that much extra revenue. It's far better (for them) to use people's expectations of such an algorithm to profit from a bait-and-switch. As an example, see Amazon's search engine.
Think they also direct listeners to cheaper to license sound-alike version of songs, especially from previous decades. I've pretty much given up on the recommendations from any of these companies. Pandora used to be reasonably good but they started playing the Studio 54 game where there was always another higher level of subscription to buy to avoid annoyances they would create. The best recommendation engine I've used was last.fm for the xbox. I could leave that on all day and rarely need to do a skip. But that was discontinued long ago, maybe ten or fifteen years ago? Haven't seen anything come close since. Amazon keeps giving me free Amazon Music but it's not even worth the bother of loading up as the system is so focused on anything and everything except my musical tastes.
Pandora has always been the best at this (for me at least). Especially because they are the only tool I know of that has a MUST HAVE feature for me: contextual thumbs up/down.
If I'm in the mood for a certain mood/genre, that's what I want to listen to. Even if a song comes on that I normally love, if it doesn't fit the mood, I want a way to say "that doesn't belong here right now". So in Pandora I start a station and can thumb down songs that don't fit that station's theme, and Pandora understands that doesn't mean I don't like that song. It just means I don't like that song in this context.
As far as I've seen, every other service only registers these globally. Either I like a song or I don't. That doesn't make any sense to me.
Finding music requires effort and energy. I rember going to the recordshop an listening to a pile of records. After 45 minutes i would have to leave and take what i found until then becaue i was exhausted. This doesnt change with music streaming. It still takes effort to discover and process and the machine cant do that for you no matter how clever the AI. Listening to new music, filtering, seeing what is in it for you is communication and as all communication basically boils down to a biological process that can reward you with dopamine or not but will require energy either way.
Here's something rather mundane, but definitely works for me: open Internet Archive's Audio section and choose "This Just In". Click on the first interesting thing you see and let it play. Most things I choose are pretty good to so-so, but sometimes I find a real gem (e.g., 3-hour-long John Peel radio captures transferred from cassette from the late 1970s).
It's all rather random but relies somewhat on your gut instinct. I find it more enjoyable than the top music streaming services. Case in point: someone uploaded an excellent field recording of a Bruce Hornsby concert from 2017 yesterday - listened to the whole thing a few times already (and I'm not really a big fan, but he's a great showman).
Try Pandora - been using it since very early days. It's the only service I've ever used that has consistently produced playlists I enjoy based often on just a couple of thumbs ups/thumbs downs.
Another good music discovery method is to use AOTY - albumoftheyear.org
I find their comprehensive section of Lists (featuring lists from all the major publications) and aggregate lists is inspiring for discovery.
For example, check out this list of the best albums so far this year according to The Quietus, containing some great stuff your algorithm would never consider:
There's a VST that creates embeddings of all the sound samples you have in a directory and then projects it down to 2d so you can visually explore and find sounds you're looking for. Similar sounds are near each other.
I always thought doing that with the entire spotify library would be amazing. Give me a low dimensional space to explore the library. Even cooler if the embeddings have similar geometric properties of language embeddings where I could do arithmetic with songs to find interesting combinations.
Really interesting idea, using images to generate playlists. I’m curious what interpretation is being done on the image.
I find myself in the spotify ‘discovery’ trap getting slowly funneled into a steady universe based on likes from previous weeks’ playlist, where everything is new but also the same. I’ve often correlated music with mood and color and it can be easier to express what you are seeking with colors vs. expressed genre, which is just a broad classification
Like many people here, I've never found algorithmic recommendations useful. What works best is what has always worked best: recommendations from people with similar taste.
Back in the day, this was friends lending me CDs. Now, I follow a bunch of people on Mastodon who almost only post about music.
I've also found following the releases from specific smaller record labels quite useful - often the artists on a label will fit a certain vibe, even if they don't specifically overlap in genre.
When I have people over, I hand them the iPad that runs the audio system (I use Roon, Qobuz and Tidal but I imagine anything will work). When they play new and interesting things, they are now in my history. My favorite discovery from this was Massive Attack's Teardrop, which was a track I had heard before (and loved) yet completely forgotten about years later.
For the electronic music sub-genre, some streaming compnay needs to buy https://www.1001tracklists.com/ for their playlist data (i.e. DJ set lists from soundcloud) and incorporate it into their recommendations. Your welcome Spotify!
I figure the problem is access. You either have the obscure music people want to discover in your db or you don't. The music available on spotify is a drop in the ocean so it doesn't really matter how searches work because a majority of the possible results simply aren't known by the platform
I don’t know if it’s the case anymore but I used to use last.fm’s radio service 8 or so years ago and it was absolutely great at finding and recommending me genuinely new stuff.
the underlying assumption for these systems for most people seems to be that all songs are treated equally.
this is clearly no longer the case for any major streaming platform. their own logic to promote might be too egregious now. same seems like for shuffling through a large playlist.
one can try to empathize to the ones designing this (e.g. shuffle anticipating network drops and switch to cached results for the next track) but self-discovery will remain evergreen.
you could suggest music from artists that have released music on the same record label within +/-5 years of what you listened to and get close enough. the human curation is already "baked in".
graph traversal playlists are the most interesting idea to me, especially if you can put some bounds on (i.e. weight positively and negatively certain artists in the graph)
On Reddit, there was a few threads where people were reporting that, no matter what song radio they listened to, they were getting Sabrina carpenter. Didn't matter if it was Rap, Sad Indie, Vaporwave; She would eventually come on.
It's noted that her current tour is sponsored by Spotify.
The service I got the most use out of was Rhapsody back in like 2005. Instead of playlists and "radio" it listened artists that influenced this band and artists that were influenced by this band. That lead directly to music I liked.
Otherwise, most of the services (Spotify, Apple Music, Youtube Music) have been really bad at recommending music. No amount of downvotes on songs seems to give their algos any input on bands I don't want to hear. Further, they always seem to devolve in to ridiculous recommendations. I've asked for "Louis Prima" radio and have them insert hip-hop or rap.
Google Play Music did better than those 3 but that died.
Unfortunately it's an extremely unprofitable industry, with very little revenue in general, and that little revenue shared by a few duopolists. Although it's an art, it's capitalistically treated like a craft, successful practitioners serving the same, safe, "good-enough", risk-free, pleasant sound again and again. I like offensive, avant-garde, creative, novel, strange music and (1) artists I love live in immense poverty (i.e. artist life) (2) even though this is the biggest passion of my in life and I think I do have some skills, working on music is 99.999...% surefire way to have financial hardship.
The setup is pretty basic. I've built a NextJS app, for the LLM model I am using open AI gpt-4-turbo and sending the images there directly without any database for images. I did a little prompting to get the same output everytime and when I get the output I make search on the Spotify API, find the songs and create the playlist with them on your own authenticated spotify account.
Likewise I also don't have a database for the emails eighter. I am using spotifys authentication
To keep your options open, it might be worth switching from GPT to either Gemini or Llama as OpenAI official policies prevent you from training on your logs with the argument this training is “illegal, harmful, or abusive,” so you’d never be able to fine tune or train your own AI in the future on your data or help others do the same.
If being permanently locked into a single intelligence service (or risking getting cut off or sued) is unacceptable for you as for me then OpenAI terms today are not acceptable. Yeah yeah maybe they won’t go after you, but why miss the opportunity for malicious compliance? Ditto Claude. Try a specialized model for your use case instead.
Gemini has no such customer noncompete, and with llama 3.1 meta removed theirs last week.
Thanks, that was something I was already considering. Do you know where can I find models specialized in this area? Because it is kinda niche, I couldn't find something that maps images into playlists directly, thats why i went for llms
I don't think he wants a passport photo, just saying any picture as a seed — just enter a photo of a dog (so long as it's not your "first pet" with a personalised name tag round her neck)
If anyone constructed a PDF, which was itself blank but, via embedded JavaScript, loaded parts of itself from a remote server, people would rightly balk and wonder what on earth the creator of this PDF was thinking — yet this is precisely the design of many “websites” [1]
Don't take it personally though, JS abuse is unfortunately common; just need to vent for once. But like the author of article I linked, people tend to just skip those blank pages. It's a tiny minority which is probably not the target audience anyway. Actual hackers would look at how you did it and make their own version.
Pandora had decent algorithms for recommending things, but it had such a small library that it would frequently repeat the same handful of albums for anything I searched for. This irks me, as I hate wearing out good music.
Spotify is currently where I keep my weeks-long playlists that I've built over the past couple decades. Even with such large playlists as input for their radio recommendations, Spotify doesn't do a very good job recommending new music either.
Whatever happened to the good databases and their algorithms? They definitely used to exist.