the problem with these algorithms is they get stuck in a loop and it keeps recommending the same type of music to you or keeping you in a bubble. Related songs to a song are different for each user so it ends up showing you more of the same (you can test this yourself, make another spotify account and check, this also applies to top songs from an artist and even spotify official playlists). At least I notice this problem with youtube music recommendations, maybe spotify is better with regards to this (I don't use spotify that much) but I still don't like it. Another problem besides getting stuck in a bubble and accidentally reinforcing certain cultures of music over others is that algorithms can be intentionally manipulated to devalue certain types of music/content and users won't notice. This gives a few tech company executives the power to shape public opinion, even to the point of changing results of elections for example, by subtly shifting what content is valued or devalued. (this also affects what content is created in the first place as creators make content that gets more views or money). This is more clearly a problem for writing/language content recommendations like twitter or tiktok style apps, but I think it could apply to music too, music is also speech and can also show people and/or normalize certain ideas.
I left Spotify when they drastically changed their UI around 2017-18, then tried out apple music but the Spotify algorithm is so hard to compete with that I went back after a couple of years. They know when you like different kinds of music and algorithm is aware of your interest in new up and coming artists and it's easy for you to basically keep the algorithm learning from you. I suggest trying it out a bit more
@RemotHuman I don't argue any of Your suppositions, how ever I would contend that these possible "traits" are not worse than other "systems" that we have had up till now. And regarding "getting caught in a bubble" have You ever listened to "radio stations", because I don't think they are really any better... In my opinion the problem is (as is often the case nowadays) rather one of scale, when You have "actors" with such an enormous footprint and reach in any "market", then what ever problem or "quirk" they have, how ever small can become a major problem for many people. Best regards
Exactly. I saw some people said youtube music recommendation is better than spotify. seriously? youtube music doesn't even recognize niche artists. I keep coming back to Spotify for the reason!
@@quas3728 YT music doesn't even change the discover mixes playlists. I've disliked several songs over and over again in my weekly discover mixes and they still pop up to this day on my recommendations. Spotify is light years ahead. Well said.
@@ptbauro It's funny that I disliked one song from ytm's official playlist and then they put music video version of that song. what's the point of dislike button when they don't use it to improve recommendation?
I switched to Apple Music a few years back when that’s all you could stream on the Apple Watch, but I still find myself going back to Spotify occasionally for its better selection of indie artists (I’m talking like Tumblr. DIY musicians that no one else has outside rough TH-cam uploads) which is odd considering Apple Music advertises over 100 million songs whereas Spotify boasts 80 million making you think the larger quantity would have the greatest selection for you, but that’s not always true. Although I will say both Apple Music and Tidal do offer high bitrates for streaming than Spotify currently does which is nice for those that have solid speaker systems at home. Klipsch reference series and the newer Sonos Arc sound bar with rear channels I feel sounds better with select albums than Spotify.
As seems to be the experience of many commentors here, algorithms/recommendations are becoming an increasingly crucial element of differentiation between Spotify & its competitors (Apple Music, TH-cam Music, Amazon Music etc). Music streaming is a unique industry in that all the players have virtually all the music content available (unlike movie streaming like Netflix where they won't have all of the world's movies), and everyone is priced the same ($10.99/month in the US). This may be how Spotify can differentiate from the crowd, though only time will tell if its enough to help drive profitability in the future. I made a video on Spotify's persistent lack of profits and how they compare vs the other streaming platforms, may be interesting to you.
I am from Chile, but Spotify assumes that we hear Mexican and Puerto Rican music here. This is like treating Canada and New Zealand as two provinces of the same country just because in both places people speak English.
I personally like Spotifys algorithm. It’s recommended so many good songs to me over the years. I’m sure it can be better, however it’s the best out there imo
Often, people get in my car with me and at some point they'll say "Ohh, I love this song! I've forgotten all about this song!" Then I answer back with something like, "That's why I still pay for SiriusXM. Even though I have Spotify Premium, you get stuck in loops on Spotify where you do hear the same songs over and over. With satellite radio, it's not trying to guess what I'll like so the variety is much wider."
Spotify is the best except for their lack of hifi quality option. Us audiophiles who want to hear all that a artist and producer out in to a track miss out and with Tidal or Apple Music for example offering it, it's a shame Spotify doesn't want to.
The largest area Spotify lags is in classical music (which admittedly probably the majority of listeners aren’t impacted by) but Apple Music’s new ‘Classical’ app is insanely good in this respect and is actually “Primephonic” which Apple bought out, and rebranded as its own if anyone remembers that from last decade. It’s able to filter and categorize even better than ‘IDAGIO’ with insanely rich ‘Dolby Atmos’ recordings. 5-star experience all the way 👍
There is a personalized DJ AI out there?! That sound awesome! I look forward to when it matures enough that I can give it input (via speech or text) to tell it what mood of music I want to listen to at any one time. Also, it would be cool if an AI can do remixes for me on the fly. A while back, I found that someone mixed Mozart with 'Like a Hurricane' and it is one of the best things I've heard.
When you upload a video, it's first only available at the lowest resolution. The higher resolutions get encoded in the background and are normally available within 30 minutes of uploading a video.
I use Apple Music because of its unparalleled music quality, but I miss Pandora’s recommendations. Discovered so many of now my favorite bands, back in the day.
0:01: 🎧 Spotify, the world's largest music streaming service, uses AI and collaborative filtering to personalize user experiences on the platform. 3:00: 🎵 Spotify's algorithm uses content-based filtering to make recommendations, but it has potential biases and limitations. 6:03: 🎧 Spotify's algorithm is not optimized for providing additional context about music, unlike Apple Music's new app. Recap by Tammy AI
As far as I can tell Spotify's algorithm functions to bring me the songs and artists I already listened to. Take a cue from Amazon's "recommended if you like" function from 20 years ago. That worked much better.
Spotify algo barely works for me as I can listen to progressive metal in the morning, phonk at work and soft jazz in the evening sporadically injecting some hip hop into it. It just does not understand what I want and tries to recommend me either of those things. But my recommendations are always terrible... It works well for people who are in a bubble and it keeps them in a bubble. The algorithm is terrible for people who have diverse musical taste. Also algo arguably tries to prioritize tracks that pay more to spotify per listen which also messes things up.
This! I would drop Spotify in a hot second if something better came along simply because the shuffle feature is an absolute dud. When I hit shuffle I want to shuffle all my songs not what it thinks I want.
This had so much potential! Interesting topic and high production value. But it didn’t really get into exactly how it worked, and instead spent 50% of the time talking about an odd bias story that isn’t really a huge concern in playlisting songs
Exactly. IKR? I already know how the algorithm works but when I found this video, I was expecting them to give us really useful Info. But turns out 80% of this video was clickbait. Now let me share some of my knowledge The algorithm will study your user habits (it will study the type of music you listen to, the artists you follow e.t.c). Once the algorithm has gotten sufficient data from you, it will now look for other Spotify users that have similar music taste to you. Any song they listen to, the algorithm recommends to you. It's as simple as that really. I really wasted my time watching this video lol
If Spotify uses AI to suggest music we have no need to worry about the supremacy of AI intellect any time soon. I’ve been using Spotify for almost 10 years and it still can’t figure out what I like because my taste is varied and not defined by genres. Spotify was the first subscription I cancelled when I need extra money because I have not been able to discover much interesting new music over that platform.
I listen to Spotify and Pandora and I put my own music that I listen to which is old time music from way back I don't listen to today's music very good info WSJ
Spotify is mostly funded by the big music labels. Meaning they can push big and new artist THEY SIGNED ON. Those music labels literally pay Spotify to show these artist as ‘trending’ wether they may be or not. Tough luck to independent artist as they can’t pay to trend. So whatever ‘AI algorithm’ Spotify is claiming to use is pretty bs
i use the AI DJ every single day and probably 50% of the time i listen to music its through the DJ. i love how it shows me stuff i love but also shows me very diverse types of music and i love it. best feature!
Is it on Spotify? or a different service? If it's on Spotify, may I ask how you use it? It sounds so cool but I am struggling finding the feature Thank youuu
I love the DJ. I've been using it since it was introduced and I found that I end up liking more songs from the DJ recommendation than other way on spotify. Yes, sometimes, it misses the mark but most often than not, I can vibe to it. From time to time you can tell its still in beta cuz it'll be like "and now we're gonna play an artist you used to listen to all the time back in the day" and I never heard of the artist. haha 2/3 times I ended up liking them tho and adding a few of their songs to my library.
This video had so much potential! Interesting topic and high production value.But it didn't really get into how it works, but instead spent 50% of time talking about an odd bias story that isn't really a big issue in playlisting songs. I already know how the algorithm works but when I found this video, I was expecting them to give us really useful Info. But turns out 80% of this video was clickbait. Now let me share some of my knowledge The algorithm will study your user habits (it will study the type of music you listen to, the artists you follow e.t.c). Once the algorithm has gotten sufficient data from you, it will now look for other Spotify users that have similar music taste to you. Any song they listen to, the algorithm recommends to you. It's as simple as that really. I wasted my time watching this lol
doesn't TH-cam technically have a bigger dataset? this is even more true for obscure songs by new artists, since TH-cam tends to recommend new stuff as well, whereas Spotify recommends things others are listening to a lot.
I highly doubt its true but if spotify are only now looking at implementing a reccomender system to provide users with song reccomendations I would be shocked. Literally one of the most commonly used algorithms in ML
I don't like Spotify because after they take their cut of my subscription fee, the remaining should be split between the artists I listen to and no others. This is important because, if, for example, I only enjoy traditional black Christian gospel music (eg Kirk Franklin), I don't really want to financially support people who overtly worship Lucifer. Or if I don't want to support a major gospel artist one month, the money, after the Spotify cut, should be given to the small startup artist I wanted to support. Patrick Riddick might have 1000 loyal fans who play his music. If we pay $10 monthly and Spotify takes 30%, a small gospel artist like Patrick Riddick should walk away with $7 x 1,000 loyal fans that month. Instead, the bulk of that $7 is maldistributed to people like Beyonce, Taylor Swift, Ed Sheeran, and Drake. The system is not only unfair to the artist, but it presents a moral conflict for conservative users. 😢
Streaming still doesn't cater to my need of having a centralised place where I have popular releases along with obscure bonus tracks or demos. I want play count tracking that is accurate as well, so I really don't like using different ways to consume music all that much.
The biggest problem with Spotify that it’s algorithms doesn’t understand who listening! For example you are traveling with your kids and play music for them. Next day your list is ready which is included Mariah Carey, Jennifer Lopez and cocomelon, baby shark songs. The second thing is that for example you are listening Turkish songs and Spotify mixed them with incompatible songs which are American pop music. It mixes two different language and culture. If it capable to make a list abstaining from these problems would be great.
Most of the recommendations from Spotify are not good for me. TH-cam usually recommends better artists for me. I guess cause most of the artists I listen to are small or not big in the west. TH-cam also has more data points, like how most touhou circles are not on Spotify
The utilization of algorithms and artificial intelligence (AI) in Spotify's platform is undeniably impressive. However, despite its advanced technological features, some experts have identified certain biases that are present within the system. It is my sincere hope that Spotify will continue to address and rectify these issues, ensuring that their platform remains accessible and equitable for all users.
Maybe they should make it clear that if you listen to podcasts then your main screen will be only podcasts. They separated the feeds, which was good, but then they turned my music feed into TikTok. How tf am I supposed to find new artists
Whatever they have done recently is not working. My "radio" is now the same songs over and over. I got SXM again because of it. I did enjoy the AI dj for a while however.
Spotify is an amazing platform for music, while its music suggestions are really really bad. 'Discover weekly' has never once amused me in the last 4 years.
While I really love spotify, the only thing over the years I have purposely avoided is listening to songs suggested by their algorithm. And it shall continue that way.
How long have you used it?? Cuz the algorithm is literally the best out there. I've found songs I never knew I would love. My current favorite songs were all discovered thanks to the algorithm
We need more of these than the "briefing" videos
Right
I agree👍
Also agree
This. Zoom calls aren’t content for a video sharing platform.
@@14x7j4 exactly right. I like to see the animation videos.
the problem with these algorithms is they get stuck in a loop and it keeps recommending the same type of music to you or keeping you in a bubble. Related songs to a song are different for each user so it ends up showing you more of the same (you can test this yourself, make another spotify account and check, this also applies to top songs from an artist and even spotify official playlists). At least I notice this problem with youtube music recommendations, maybe spotify is better with regards to this (I don't use spotify that much) but I still don't like it.
Another problem besides getting stuck in a bubble and accidentally reinforcing certain cultures of music over others is that algorithms can be intentionally manipulated to devalue certain types of music/content and users won't notice. This gives a few tech company executives the power to shape public opinion, even to the point of changing results of elections for example, by subtly shifting what content is valued or devalued. (this also affects what content is created in the first place as creators make content that gets more views or money). This is more clearly a problem for writing/language content recommendations like twitter or tiktok style apps, but I think it could apply to music too, music is also speech and can also show people and/or normalize certain ideas.
Amen
They address that around 4:30 in the video.
I left Spotify when they drastically changed their UI around 2017-18, then tried out apple music but the Spotify algorithm is so hard to compete with that I went back after a couple of years. They know when you like different kinds of music and algorithm is aware of your interest in new up and coming artists and it's easy for you to basically keep the algorithm learning from you. I suggest trying it out a bit more
@RemotHuman
I don't argue any of Your suppositions, how ever I would contend that these possible "traits" are not worse than other "systems" that we have had up till now. And regarding "getting caught in a bubble" have You ever listened to "radio stations", because I don't think they are really any better...
In my opinion the problem is (as is often the case nowadays) rather one of scale, when You have "actors" with such an enormous footprint and reach in any "market", then what ever problem or "quirk" they have, how ever small can become a major problem for many people.
Best regards
Do you have a degree in any tech related fields? If no, then you can’t understand.
I’m surprised that with the advanced algorithm and massive database, Spotify so rarely suggests a new song that I actually like.
You sure you're using Spotify? Cuz mine recommends songs I like everytime
@@Tobby_ Yep.
Yeah it just keeps playing the same songs over and over again. I'm reconsidering Pandora
I've tried every major streaming service... Spotify's algorithms are light years ahead of everyone. That + curated playlist = no competition.
Exactly. I saw some people said youtube music recommendation is better than spotify. seriously? youtube music doesn't even recognize niche artists. I keep coming back to Spotify for the reason!
@@quas3728 YT music doesn't even change the discover mixes playlists. I've disliked several songs over and over again in my weekly discover mixes and they still pop up to this day on my recommendations.
Spotify is light years ahead. Well said.
@@ptbauro It's funny that I disliked one song from ytm's official playlist and then they put music video version of that song. what's the point of dislike button when they don't use it to improve recommendation?
I switched to Apple Music a few years back when that’s all you could stream on the Apple Watch, but I still find myself going back to Spotify occasionally for its better selection of indie artists (I’m talking like Tumblr. DIY musicians that no one else has outside rough TH-cam uploads) which is odd considering Apple Music advertises over 100 million songs whereas Spotify boasts 80 million making you think the larger quantity would have the greatest selection for you, but that’s not always true. Although I will say both Apple Music and Tidal do offer high bitrates for streaming than Spotify currently does which is nice for those that have solid speaker systems at home. Klipsch reference series and the newer Sonos Arc sound bar with rear channels I feel sounds better with select albums than Spotify.
It’s literally not even Spotify, you’re thanking them for using your data to progressively take away your desire to choose your own music 😂
As seems to be the experience of many commentors here, algorithms/recommendations are becoming an increasingly crucial element of differentiation between Spotify & its competitors (Apple Music, TH-cam Music, Amazon Music etc). Music streaming is a unique industry in that all the players have virtually all the music content available (unlike movie streaming like Netflix where they won't have all of the world's movies), and everyone is priced the same ($10.99/month in the US). This may be how Spotify can differentiate from the crowd, though only time will tell if its enough to help drive profitability in the future. I made a video on Spotify's persistent lack of profits and how they compare vs the other streaming platforms, may be interesting to you.
I am from Chile, but Spotify assumes that we hear Mexican and Puerto Rican music here. This is like treating Canada and New Zealand as two provinces of the same country just because in both places people speak English.
I personally like Spotifys algorithm. It’s recommended so many good songs to me over the years. I’m sure it can be better, however it’s the best out there imo
spotify needs to introduce an option to give negative feedback on my home screen recommendations
Just make the shuffle actually shuffle. I keep hearing the same 5 songs on my 300 song playlist
Often, people get in my car with me and at some point they'll say "Ohh, I love this song! I've forgotten all about this song!" Then I answer back with something like, "That's why I still pay for SiriusXM. Even though I have Spotify Premium, you get stuck in loops on Spotify where you do hear the same songs over and over. With satellite radio, it's not trying to guess what I'll like so the variety is much wider."
Spotify recommended a playlist way better than m own. I'm usually picky with music, but have to admit they did a great job
Spotify Algorithm = play the same 17 songs in a genre for eternity
exactly!!!
Spotify is the best except for their lack of hifi quality option. Us audiophiles who want to hear all that a artist and producer out in to a track miss out and with Tidal or Apple Music for example offering it, it's a shame Spotify doesn't want to.
It literally uses 320kpbs of Opus audio. Touch some grass.
@@aaronwestley3239Last time I checked, it uses Vorbis not opus
To be honest I think TH-cam does a much better job of serving me new interesting music mixes than any of these streaming services
The largest area Spotify lags is in classical music (which admittedly probably the majority of listeners aren’t impacted by) but Apple Music’s new ‘Classical’ app is insanely good in this respect and is actually “Primephonic” which Apple bought out, and rebranded as its own if anyone remembers that from last decade. It’s able to filter and categorize even better than ‘IDAGIO’ with insanely rich ‘Dolby Atmos’ recordings. 5-star experience all the way 👍
There is a personalized DJ AI out there?! That sound awesome!
I look forward to when it matures enough that I can give it input (via speech or text) to tell it what mood of music I want to listen to at any one time. Also, it would be cool if an AI can do remixes for me on the fly. A while back, I found that someone mixed Mozart with 'Like a Hurricane' and it is one of the best things I've heard.
How is 360p the maximum resolution?
When you upload a video, it's first only available at the lowest resolution. The higher resolutions get encoded in the background and are normally available within 30 minutes of uploading a video.
I use Apple Music because of its unparalleled music quality, but I miss Pandora’s recommendations. Discovered so many of now my favorite bands, back in the day.
Very fancy, but when are they fixing the shuffle mode in Liked Songs?
AI is like Teflon, once it’s in the environment it will be impossible to remove.
Yh hence why we need it to enter asap so it gets embedded
0:01: 🎧 Spotify, the world's largest music streaming service, uses AI and collaborative filtering to personalize user experiences on the platform.
3:00: 🎵 Spotify's algorithm uses content-based filtering to make recommendations, but it has potential biases and limitations.
6:03: 🎧 Spotify's algorithm is not optimized for providing additional context about music, unlike Apple Music's new app.
Recap by Tammy AI
Many points exploring, good WSJ.
The shuffle sequence for my playlists is the same most of the time, that sucs.
WSJ Official just posted 360p????
As far as I can tell Spotify's algorithm functions to bring me the songs and artists I already listened to.
Take a cue from Amazon's "recommended if you like" function from 20 years ago. That worked much better.
Thanks for reminding me of "Black and Yellow".
I think that Spotify's ability to store and play offline is outstanding compared to TH-cam.
I only used Spotify, but when used TH-cam Music or Pandora, I can clearly see they are above Spotify in terms of music recommendations.
Amazing video. Thank you
Ok that's nice, now please give us a TRUE shuffle button that's actually random and not based on algorithms. It's just patronizing
Spotify is an AI genius. That's why I have always loved it!!! Amazing to see this at work here.
I love theses types of videos
Spotify algo barely works for me as I can listen to progressive metal in the morning, phonk at work and soft jazz in the evening sporadically injecting some hip hop into it. It just does not understand what I want and tries to recommend me either of those things. But my recommendations are always terrible...
It works well for people who are in a bubble and it keeps them in a bubble. The algorithm is terrible for people who have diverse musical taste.
Also algo arguably tries to prioritize tracks that pay more to spotify per listen which also messes things up.
This! I would drop Spotify in a hot second if something better came along simply because the shuffle feature is an absolute dud. When I hit shuffle I want to shuffle all my songs not what it thinks I want.
This had so much potential! Interesting topic and high production value. But it didn’t really get into exactly how it worked, and instead spent 50% of the time talking about an odd bias story that isn’t really a huge concern in playlisting songs
Exactly. IKR? I already know how the algorithm works but when I found this video, I was expecting them to give us really useful Info. But turns out 80% of this video was clickbait.
Now let me share some of my knowledge
The algorithm will study your user habits (it will study the type of music you listen to, the artists you follow e.t.c). Once the algorithm has gotten sufficient data from you, it will now look for other Spotify users that have similar music taste to you. Any song they listen to, the algorithm recommends to you. It's as simple as that really. I really wasted my time watching this video lol
it is not the bias that is what we should worry, it is the dependency to technology! it is wierd that it looks like from the movie "Her" that DJ AI.
If Spotify uses AI to suggest music we have no need to worry about the supremacy of AI intellect any time soon. I’ve been using Spotify for almost 10 years and it still can’t figure out what I like because my taste is varied and not defined by genres. Spotify was the first subscription I cancelled when I need extra money because I have not been able to discover much interesting new music over that platform.
I listen to Spotify and Pandora and I put my own music that I listen to which is old time music from way back I don't listen to today's music very good info WSJ
360p??? really?
Spotify is mostly funded by the big music labels. Meaning they can push big and new artist THEY SIGNED ON. Those music labels literally pay Spotify to show these artist as ‘trending’ wether they may be or not.
Tough luck to independent artist as they can’t pay to trend. So whatever ‘AI algorithm’ Spotify is claiming to use is pretty bs
people dont talk about this enough
i use the AI DJ every single day and probably 50% of the time i listen to music its through the DJ. i love how it shows me stuff i love but also shows me very diverse types of music and i love it. best feature!
Is it on Spotify? or a different service?
If it's on Spotify, may I ask how you use it?
It sounds so cool but I am struggling finding the feature
Thank youuu
@@r6m697 premium only in us canada uk and ireland
I love the DJ. I've been using it since it was introduced and I found that I end up liking more songs from the DJ recommendation than other way on spotify. Yes, sometimes, it misses the mark but most often than not, I can vibe to it. From time to time you can tell its still in beta cuz it'll be like "and now we're gonna play an artist you used to listen to all the time back in the day" and I never heard of the artist. haha 2/3 times I ended up liking them tho and adding a few of their songs to my library.
360p is the max? Or is it just me?
Just you
So theyre basically just using k-nearest neighbor alg :P
Jordan Kranse is a massive Jonas Brothers fan and I love that
The DJ is one of my boyfriends favorite features. It needs improvement in terms of recommendations and variety but that will come in time
why is 360p the max?
helpful video for a dev like me
This video had so much potential!
Interesting topic and high production value.But it didn't really get into how it works, but instead spent 50% of time talking about an odd bias story that isn't really a big issue in playlisting songs.
I already know how the algorithm works but when I found this video, I was expecting them to give us really useful Info. But turns out 80% of this video was clickbait.
Now let me share some of my knowledge
The algorithm will study your user habits (it will study the type of music you listen to, the artists you follow e.t.c). Once the algorithm has gotten sufficient data from you, it will now look for other Spotify users that have similar music taste to you. Any song they listen to, the algorithm recommends to you. It's as simple as that really. I wasted my time watching this lol
Tiktok algorithm next please
Why is this video in 360p?
im music person, using apple music more convenient rather than spotify. but it still bug everywhere. mainly in mac version.
Makes sense, the same songs I use to love I now hate because of this very algorithm. They play the same songs OVER AND OVER!
But they can’t make a basic shuffle that my CD player in the 90s could manage
Amazing content. Thank you guys. 🎉
doesn't TH-cam technically have a bigger dataset? this is even more true for obscure songs by new artists, since TH-cam tends to recommend new stuff as well, whereas Spotify recommends things others are listening to a lot.
their algorithm is the reason I can't go away and use another streaming service... just the best out of everyone
If only this worked. I’ve tried this new Dj X and it’s always trying to push music I would never even consider.
I highly doubt its true but if spotify are only now looking at implementing a reccomender system to provide users with song reccomendations I would be shocked. Literally one of the most commonly used algorithms in ML
The VP sounds more AI than the AI DJ😂
I don't like Spotify because after they take their cut of my subscription fee, the remaining should be split between the artists I listen to and no others. This is important because, if, for example, I only enjoy traditional black Christian gospel music (eg Kirk Franklin), I don't really want to financially support people who overtly worship Lucifer.
Or if I don't want to support a major gospel artist one month, the money, after the Spotify cut, should be given to the small startup artist I wanted to support.
Patrick Riddick might have 1000 loyal fans who play his music. If we pay $10 monthly and Spotify takes 30%, a small gospel artist like Patrick Riddick should walk away with $7 x 1,000 loyal fans that month. Instead, the bulk of that $7 is maldistributed to people like Beyonce, Taylor Swift, Ed Sheeran, and Drake.
The system is not only unfair to the artist, but it presents a moral conflict for conservative users. 😢
360p，在搞什么啊
Amazing video, I'm a proud WSJ subscriber😊
It all comes back to linked data. Do an explainer on that and we’ll all know more about all of this tech world we live in
TH-cam will forever be my spotify
I trust Spotify A LOT more than Apple though like A LOT
Streaming still doesn't cater to my need of having a centralised place where I have popular releases along with obscure bonus tracks or demos. I want play count tracking that is accurate as well, so I really don't like using different ways to consume music all that much.
still i think suggestions by yt music are the best. You have to play just one song and here you go
the same happens with podcasts, of course, which is even more dangerous than the biased music implies
Been waiting for this AI DJ in Europe, but still haven't got it
The robot hasn't figured me out yet 🥴
Garmin and Coros seem to have the fitness space all locked up, well, the ones not controlled by Apple.
Polar and Sunto seem to be giving up
there aint gonna be much tranparency no matter how much you try to improve the algorithm if payola still exists
The biggest problem with Spotify that it’s algorithms doesn’t understand who listening! For example you are traveling with your kids and play music for them. Next day your list is ready which is included Mariah Carey, Jennifer Lopez and cocomelon, baby shark songs.
The second thing is that for example you are listening Turkish songs and Spotify mixed them with incompatible songs which are American pop music. It mixes two different language and culture. If it capable to make a list abstaining from these problems would be great.
Most of the recommendations from Spotify are not good for me.
TH-cam usually recommends better artists for me. I guess cause most of the artists I listen to are small or not big in the west.
TH-cam also has more data points, like how most touhou circles are not on Spotify
Spotify algorithm is nothing to brag about, it’s terrible.
Watching from India sir
The utilization of algorithms and artificial intelligence (AI) in Spotify's platform is undeniably impressive. However, despite its advanced technological features, some experts have identified certain biases that are present within the system. It is my sincere hope that Spotify will continue to address and rectify these issues, ensuring that their platform remains accessible and equitable for all users.
ChatGPT?
@@skyplayx3558 yes brother i just prompt chat gpt to make my comment more formal 😇
Maybe they should make it clear that if you listen to podcasts then your main screen will be only podcasts. They separated the feeds, which was good, but then they turned my music feed into TikTok. How tf am I supposed to find new artists
購買意欲を高めるための広告に音楽要素が加わると、ノリの加速度の付き方が格段に違ってくる。
個人的に危険だと感じるので、タイプの違うBGMを複数用意しとく必要があり、You Tubeと併用しているが、選曲に違いが出にくくなる。
Spotifyは、もともとGoogleMailやMapの入力内容にあわせたBGM流すのが興味深かった。
But where is true Shuffle
They really need to fix those tracks that say an artist is involved in it, when in truth they have nothing to do with it.
6 degrees of kevin bacon has been replaced by 6 degrees from 7 nation army.
Spotify is the one of the subscription that i use the most
Whatever they have done recently is not working. My "radio" is now the same songs over and over. I got SXM again because of it. I did enjoy the AI dj for a while however.
5:00 Bias against South Asian music
Spotify is an amazing platform for music, while its music suggestions are really really bad.
'Discover weekly' has never once amused me in the last 4 years.
360p?
While I really love spotify, the only thing over the years I have purposely avoided is listening to songs suggested by their algorithm. And it shall continue that way.
Wish they’d pay the artist what their streams are worth 5 cent per stream is sickening
I've used Spotify for years and never used recommended playlists. I rely more playlists created by regular users with hand-picked songs.
Thank you
It doesnt work, spotify cnstantly brings up songs that I always skip again and again
Ever heard of resso??
This video was expensive 😢😂
AJ Rafael!!
Cool
📣📣📣NOT FORGET THE LOST ART OF JUST WALKING IN TO YOUR LOCALLY-OWNED RECORD STORE❗❗
*broke android noise intensifies*
At a top 5 percent listener I have to admit the algorithm is terrible. I constantly skip tracks and am often sickened by suggested tracks
How long have you used it?? Cuz the algorithm is literally the best out there. I've found songs I never knew I would love. My current favorite songs were all discovered thanks to the algorithm
Wait... they're not using reinforcement learning? Interesting.
I would happily pay for a human-curated streaming service.
Bandcamp have human recommendations