Joe Mills (Aver) of Move 78 decodes jazz from the algorithms

LP: Good afternoon, good morning, whatever it is where you are. 

Joe Mills: Evening-ish. Yeah, early evening. 

LP: Thanks for making time. I very much appreciate it. 

Joe Mills: Well, good man. Thank you for having me on. 

LP: I'm very excited to talk with you. It's interesting the way I came across your music. You may or may not be interested to know it was the Spotify algorithm.

Joe Mills: I'm intrigued to know constantly because it also plays into a little, not a trick, but a little joke hidden in the album title. So yeah. 

LP: I'm a big partisan of Release Radar on Spotify. It's a source for me for new music. And you popped up in there. But the interesting thing to take a step back about that, not specifically as it relates to you and your music, it's very interesting how I have found over the years you can train the algorithm if you put intention into it, as a listener even, you know? If I've decided, I've had enough instrumental ambient music for a while. I will spend the next two weeks listening only to old-school hip-hop and then watch my recommendations change. I guess it's working as advertised, but it's fascinating. It's fascinating.

Joe Mills: It's a funny thing that my friends comment on if they have a different music taste to me, and then I commandeer their Spotify for a bit, and then they're like. Two weeks later, I kept getting this stuff. Like why, this is your fault. Because, obviously, I've put in some Art Ensemble of Chicago or some weird Herbie Hancock or whatever that mixes in with their ambient house music that throws it all off. I like messing with people's algorithms in that way, as well. So it's good. 

LP: I think, outside of the music itself, one of the things that I think is self-evident or is obvious in what you're up to. And what Move 78 is up to is it seems to be this dialogue around an ambivalence with technology, if I'm reading it correctly. 

Joe Mills: Yeah. 

LP: And I wonder, for the benefit of our listeners, I think we need the origin story of the band name and how that relates to your sort of discourse and thinking about technology and our relationship to it. 

Joe Mills: Yeah, so for anyone who doesn't know, it is related to a famous match. It's a Go match.

So Go is an ancient Chinese board game. Google acquired a company called Deep Dream, which was creating an algorithm to play the game Go. The algorithm is called AlphaGo, and there's a documentary that showcases this. It's very similar to Kasparov versus Deep Blue, a similar man versus machine, but involving chess.

That was a massive promotional scheme for IBM. There's a huge controversy about whether or not they rigged it. It's neither here nor there, but this was a similar promotional drive in some way by Google for their AI potential.

At first, they played some other human Go players, AlphaGo, the algorithm, and it destroyed them. And then they put it against basically the Michael Schumacher or Pete Sampras of Go. This man, Lee Sedol, was the greatest player on earth at that point. And everyone thought that Lee Sedol would outsmart the computer. There was no way it was up to speed with him so far. In the five-game match, the first three games, AlphaGo completely destroyed Lee Sedol and won them outright. And then there was even commentary from the experts within the game being so astounded at how beautiful it was.

AlphaGo was playing the game, and then in game four, LisaDoll, having already lost the match technically, began to change how he played. He adapted his creative, intuitive, and skillful approach to playing the game of Go to how the computer played it. At some point, he played a game like midway through; he played a move midway through the game, which was so complicated that it baffled the computer.

In the documentary, all the scientists behind the screens say, " Oh, what just happened? All the probability rates drop, and everyone gets confused. And then, there is a human who is playing AlphaGo's moves for him. When he plays the next move that AlphaGo has selected, he looks embarrassed, and the computer starts to malfunction.

It loses the game, and then this Lee Sedol is adored, walks into a room where people are clapping hands rapturously and essentially is referred to as this God move or something so complicated it confused the computer. This move was "move 78." 

There is a more concise description of this on the inside of our first album cover, but essentially, it's the modern version of man versus machine. So not Terminator, not like, Oh, it's going to kill us. It's going to exterminate us. It outsmarted him in such a way, but the way Lee Sedol refers to it is that he learned to play Go better by playing the computer. And this is then tying it to the specific nature of how we make our music.

I am a hip-hop producer who cannot play a singular note. I do not have any formal musical training in terms of theory. I can probably show you where middle C is, but I can't play chords. And then everyone else in the band is either a classically trained jazz musician or a lifelong jazz and groove aficionado. Some people are members of the Berlin Philharmonic Orchestra.

Originally, I sat at home with my then partner and had this idea, like, wouldn't it be cool if you put at the center of a band this idea that humans react to technology? Because we're drum and bass music, or breakcore, or glitchy hip hop. Can they replay it? And also, do they play differently if they try to replay it?

And so initially, this started as a set of trials that we did, where they replayed some of my music, but then I actually, the moment when I knew it was going to work, where I played them weird clusters of sounds that I hadn't fully structured into songs yet. They improvised for 20 minutes around it.

Having previously only ever used records, free jazz, normal jazz, psychedelic rock, or whatever, you're left with just one stem. And you could get in the free jazz, all these different little parts you can piece together. But here, I would present it with 20 stems or, for the listener, the uninformed listener, 20 audio tracks comprising the one song I could pull apart and rearrange.

And then, I would take the songs back to the musicians and get them to improvise around their recordings, feeding into the deep learning idea. You learn from your own, so AlphaGo learned from its mistakes in games it played theoretically.

And this is the musicians playing to their own improvisations and then going further into the grooves. You can read it as our retaliation to technology. Still, it's more a learning from computer programs, algorithmic synthesis with granular synthesis, or, yeah, granulation, which we use.

We're using more in future albums where you're taking computer algorithms to select these random pieces of sounds and then rearrange them, and then playing around that. And again, it's this challenging the quite talented musicians to play outside the normal realms in which they would play. So yeah, it's quite convoluted in some ways, but hopefully, the music sounds good.

Suppose you choose to look into what the name means and what the. The second album is called Automated Improvisation. That is an exact description of the music. It is hours upon hours of improvisation of them being automated by me and then cycled over and over again. 

LP: All right. There's a lot there. So let me … 

Joe Mills: Sorry, once it goes, it goes.

LP: No, it's a lot. And it's a lot to explore. And it's a lot that I'm fascinated with. Let's take everything you and I just talked about for the first 10 minutes, set it aside for one moment, and say, as a listener, that's largely irrelevant. The process is largely irrelevant. This is not necessarily process music as someone would think of early minimalist music, or it's not; it doesn't sound like game music.

This sounds like a modern acoustic, electronic trip-hop, and people will recognize this as modern music. This is not a gimmick. It's just that you have a process you've uncovered. that allows you and your collaborators to access creativity differently.

Joe Mills: Yeah, essentially in my head with art in particular, not saying this is art, but the way I would approach it is I want to look at a painting and be like, that's interesting. And if there's a sub-story or all these little hidden artifacts within it, then I find out more about the artist, the time it was made, and everything else involved in the picture from that.

Then, that's my choice to dive in deeper. But my initial thing is I want to look at it and that it will be stunning in some way. And this is the kind of thing we want in the music. I don't want it to be too wanky. I don't want it to be 25-minute solos for every band member. I want it to be enjoyable on both the surface level.

And if you choose to dive deeper, then essentially … so, yeah, forget the first 10 minutes. I hope you enjoyed it. 

LP: Take care, bye. (laughter)

I must re-enter some of that stuff for a few moments and dive in and out. How did you become familiar with Lee Sedol's story, and how did it intersect with the path you were on anyway?

Joe Mills: Yeah, it's also tied into the first band name, the first album, The Algorithm Smiles Upon You. There are several coincidences that occur. So, I've been making hip-hop music since I was like 17. Cutting up samples, drum breaks, whatever, you know. My friends and I would do that for fun. And at some point, we started doing shows, and then I did a degree in design so I could. Do all the trimmings around the artwork. Cause I love vinyl covers as well as videos.

And so we put out our first record in 2011. It never took off how I wanted it to, with various complications within the band and friendships and just making obscure music that people didn't give a shit about. And then, I moved to Berlin in 2016 on the back of Brexit. By this point, I'm an avid fan of DJ Shadow and making instrumental music.

I was trying to layer it and make it atmospheric so that, in its own right, it's listenable. And so this is then the deep listening of incredibly weird music, concrete or crazy synthesis records from the sixties and seventies, free jazz, all these kinds of slightly impenetrable records trying to find sample sources that hadn't already been used by everyone that comes before you.

Yeah, this, you find me then in Berlin, where I decided to do music 40 hours a week to train and get much better. And I moved here to practice. At this exact point, my sister, who had moved to Berlin with me, she's quite a handsome lady. She'd never used a dating app before. And she used Tinder once when we'd moved to the new city, and she went on one date with one guy. This guy is now essentially sort of my brother. He is Nir Sabag, who is the drummer in the band. She met up here off of him famously. And then she said, Oh yeah, I went on a date with this guy, and he's a jazz drummer.

And I was like, Oh, all right. Okay. So what's the jazz world saying? Because. At this point in Berlin, it was very technocentric. We moved in the winter, so it was lots of clubs, late nights, and listening to music I don't particularly like. Then it was like, all right, the jazz world here is very much underappreciated.

As I said earlier, I began hanging out with many musicians. They've got terrible taste in album covers. They've got terrible taste in aesthetics in terms of the way they present their music. And so I was in that exact period of time thinking, this would be cool if we could get that together. I then watched the AlphaGo film, in bed with my then-partner. And I remember saying to her like, yo, that, that moment is so cool. It's cause everyone's so shocked by how he plays the move. He loses the game eventually.

This is the other, the real sad thing of it. It's not like I think we will outsmart the computers or anything. He is essentially doomed to be outsmarted by the computer. 

LP: And walks away from the game ultimately. 

Joe Mills: Yeah. Yeah. Well, he lost the fifth game. He said he's played AlphaGo since and continues to lose to it. But at this exact point, I would have been signed to Village Live Records, a small label in the UK that has released one instrumental album and started work on the second one. I recruited Nir to play drum breaks for me, and I recruited Doron, who was the keyboard player, to play keys on some of the tunes. I was getting frustrated with the limitations of samples. And so, to try and expand this, I'd never recruited people other than my older brother. That's the exact point when I watched the film, and Daron and Nir asked me if I wanted to start a band with them.

And I was like, yeah, I was like honored in some way. I guess I had imposter syndrome hanging around with these people who played their instruments since they were four years old. Yeah, there was a little bit of, what will I do? What do I bring here? So this is like 2018. By this point, the thing starts to swirl together. Also, I wanted to have the word move in the same way, like Rage Against The Machine, people refer to them as Rage. 

And also with the band Can, as opposed to can't, you can or move. That was a good thing to have, and then, Oh yeah, but wouldn't it be cool if you had that idea with the man versus machine thing at the center of it? And that's then when it started to swirl a bit more. So, like 2018, the game took place in 2016.

The documentary is in late 2016, 2000, early 2017. And then, about a year later, I've become aware of it and never heard of Lee Sedol before. I never even heard of Go. I was not aware for all my sins. 

LP: Does he know about you?

Joe Mills: This is one of the nerdiest but most wonderful things. I've got a friend who is a producer called DJ Food, Strictly Kev.

He shared one of our records, and I got a message from a man; his artist name and Instagram handle is Kid Vector, but he is the head of design for DeepMind. And he said, yo, I saw the album cover, I listened to the music, I love trip-hop. I love that you turned the dots of the positions into the cover.

And then I've since sent him a care package with the records and so on. But I don't think Lee Sedol, even though his voice is featured on one of the records, is aware of who we are. Not yet, anyway.

LP: I love the album art. I love the various takes on the move and the board layout. It was especially fun because I didn't know what it was at first. Like you, I've heard about Go over the years. As a technology enthusiast, I've known it's this massively complex game for computer scientists to model against. It's such great album art; it's so evocative. It works really well. 

Joe Mills: But this is another one of these things where because again, I did design, but all the artwork is done by my good friend James Brooks, and I remember that I'd done like this proper crude mock-up of like, wouldn't it just be great to have the dots on the cover?

And I remember people being like, that's just well boring. And I was like, no, but if you, if you make it, if you design it real nice and you get the lines and the form and so on. And he and I sat in a pub in Neukölln, and there's one even better bit. So, on the first record we ever did, it's like a transparent disc.

And I was saying, Oh, wouldn't it be killer if you got it so that the dots, for anyone else that's seen the artwork, this is going to make sense. But if you got it so that the middle dot, the spiral goes through on the record player, if that dot was move 78, so when you slide it on, no one else would know.

But if you were looking like, why is it this? And he completely ignored this, and then about five minutes later, he was like, I got it! We should do this thing where you put the thing in the middle. It's like these ideas; they're all sat around slowly. Like, we had a similar thing with a song title recently.

Everyone is now thinking in the same way, in this sort of computer, man versus machine, what does it represent, and what do the songs mean, and yeah, so everyone's moved into the gear of having fun with it. But he makes it the slick wonder that it is like he realizes my design dreams in such a way that I couldn't, basically.

LP: Something you said a few minutes ago, I wanted to ask you again about when you wear your producer hat in this type of music. The producer and composer, to be reductive, are essentially synonymous, right? Like the beat maker or the music bed maker. You were curating the samples and the portions of the improvisations that you would then present back to the players to respond to.

Are you looking at leveraging technology to hand that role over to, or at least to create the initial you're going to let the machine start to give you the things to react to? 

Joe Mills: Right. No, but this is... I appreciate that you and I can venture off into the nerdiness, but whether or not your listeners … 

LP: Oh, we're very process-driven here. So people love it. 

Joe Mills: Yeah. Yeah. I was talking about this thing called granular synthesis, which was developed by a guy named Robert Henke. And it's predominantly used in sci-fi soundtracks to create weird, disgruntled sounds in some way. But Daron, the more technically minded member, at the start, when I was trying to get everything together, was very prominent in prodding me to use weird rhythms, patterns, and granular synthesis. And what it does is when you normally sample something like I'm just going to show you on screen, but if you go A to B, it just goes like this, and then you chop the size in a very linear, normal way.

And what granular synthesis does is take loads of tiny little fragments of sounds, which you can determine the size of, and then rearrange them. So one of the examples we got is an unused flute solo on the tune Daisies when we play live; because it's in key with the tune, I then use the granulator to play a solo live.

The best feedback we've got so far on stage is that people can't tell what's going on. And so you've got this then flute solo that initially starts like, but if you turn the grains smaller, it then becomes this sort of distorted glitched-out sound. You can turn that into either a disturbing thing or a really beautiful thing, and you can take up the pitch and approach it in a way where the computer is doing some of the work, but you're controlling it, if that makes sense.

But it is creating such a randomness that it's not something that I decide. I'm twiddling two or three controls, but there are certain points. The best example is on track five on the second album, Automated Improvisation. The song is called Ultra Natural. We had a horn session where, for some reason, we ended up with three saxophone players in the studio, and it was a nightmare.

And there was only one usable bit of sound, which was this swell, where they all played at the same time. And then Daron sat at his house, twiddling with some knobs, and at some point, hit the right parameters, and when he pressed on the key, it just unleashed this sort of orchestral soundscape.

And then, if he played more keys up and down, it created this whole other world. And he and I sat there laughing because he was essentially holding one key down, and this whole other thing unfolded. So, to answer your question, we're not surrendering it. But we're like lending it artistic license to wander off where it wants.

And so on stage, I now place that on top of the music. For the live performance, we also like self-sampling on stage and trying to avoid getting too strange with it, but have it that the songs also go in completely different directions every night. So they might start with the bass line in the same place, but invariably, they will tangentially end up somewhere strange.

LP: I'm fascinated by how this approach extends to your live performances and into the next iteration of your studio work. It's fascinating to read about your process, so I wonder if I could start — or if you could help me start — maybe at the genesis of a track. You and your collaborators go into the studio, and it's essentially a series of jam sessions. Is that correct? Or structured improvisations? 

Joe Mills: Yeah, pretty much. So, the third album is coming out on the 3rd of November, called Grains, but we just recorded the fourth album or the initial stages of the fourth album. A good example is in our second session ever, and they played a song we liked.

It didn't quite hit, but I made a version of it. Long story short, it could have been better. And then, in this new session, in the fourth session, I took the horn lines and the string parts that we had recorded to their thing and then played the track again. So they start playing their own bass line, they start playing their own drum shuffles, they start playing their own key parts, and then granulating this horn sound on top of it and playing the strings.

And putting vocal samples on and whatever, even though they're playing a song they'd already played, they played it completely differently, and it ends up being 15 minutes long. It's much looser and weirder regardless of whether it was something that they played or not. They groove around it so that within two minutes, it's completely, although the chords might be the same, it's completely different from how it was originally presented to them because, for instance.

If Daron is in a bad mood, he plays more moodily, so the chords are sharper or longer, or there's more delay on them, or there's something more aggressive about it. And then the rhythm section played together so frequently, the bass and the drums, they also provoke Daron in a funny way. So what you'll get in the actual session is them toying around with him and him, just this kind of interplay when they play so frequently together, with me noodling around on top in some way or another.

So that's like the starting point. Then now the raw product that I'm left with, I've got 15 minutes of bass, both DI'd and with a mic in an amp out in the room, three keyboards with various pedals attached to them, me on the sampler, and then ten mics on the drum kit. And then after that. I've started to chop this up, and it will end up being reduced into parts that I think have some novel interest, for lack of a better term.

Also, we've got a song called Follow the Earworm. A lot of it is this if I've got this thing stuck in my head because I've also now been, we did this recording session three weeks ago, four weeks ago, I'd wander around the city on the trains and my bike, listening to this stuff constantly so that I can find where the parts are and then.

I already make a mental map about which bits are valuable and which will go good next to each other. And so then started assembling that. Then, some jazz aficionados will be slightly perturbed by putting things on grid in one way or another. So, nothing is quantized directly.

There is no machine-like snapping everything in place. But what I would do is, very often, with the bass and the drums, there's these, say, 32 bars that I love. For example, I would take these bars; if the rough BPM is 85 BPM, I would cut it so that the same iteration between the bass and the kick is maintained.

Because often, the bassist will play shortly after the kick, not directly on. And so this sort of interplay is maintained. But then it makes the structure malleable. So if it's on a grid in some way, then I can take, or this snare doesn't hit, I can take this snare and put this one here, or actually I want this to be quieter, so I move it around, then I will tend to the keys, then I'll tend to my part, and at this stage, we then get to.

What is probably at the end of phase one, I will then take this to Meravi, who is the French horn player, or A.J.Nilles, who is the viola player, who's in the Philharmonic Orchestra, and various other people, and we then let them improvise on top. And in these sessions, I'll have come up with a line in my head where I'll be like, oh, I want it to go (sings some notes).

And then Doran will translate that into actual musical notation. And then Meravi will play it, but she will also improvise around it. And then again, I take this away, cut it up, and work out the bits. Also, when we're recording, I pretty much know the bits that are right. The most layers it's had is probably like 10.

Six or seven additional recording sessions without a percussion or something else. And chopping up and resampling and so on. But often, it's more like the initial recording session, then one more recording session, then a resampling session. Then you get into about a three or four-minute tune — the final tune on automated improvisation, which is nine and a half minutes.

I worked out there were over 120 minutes of raw material in the session that I filtered down. Yeah. It's a long-winded process if you start to think about it in terms of time. But I have this thing where people have asked me, like, why would you do that? Why don't you just get them to play it? My friend said this thing: I've gone as far as you can go to try and not learn how to play an instrument. But I love the puzzle part of it, but also, it's leading to this nice end product where it doesn't feel forced. There's some innate strangeness to it that I like.

LP: There are many elements in what you're doing and the way you approach it that are very much analogous to a lot of mathematical theory, a lot of computer science theory, the notion of like recursiveness, and even there's like a fractal nature to what you do. It's very interesting.

And also the idea that if you had a hundred and eighty-minute improvisation, jam session, or session of material, it's very much like the game of Go in that the potential permutations and moves out of all that. The amount of material that can be spawned from one of your sessions is immeasurable.

Joe Mills: This is actually how we've come to have the first album. I'd been working on the nine tracks for Automated Improvisation, and then the pandemic hit, and we released Middling. Well, no, we hadn't released Middling. I'd sent it to the label and said I've got this; these bits sat around. I don't think they're going to go on the main album.

And he was like, Oh, we should put it out. And we would get 50 copies, lathe cut by this guy I knew in Manchester. And then we put that online, and then the distributor waved his magic wand, and it ended up on two playlists with about 2 million monthly listeners. And then, from this point, the stats went crazy.

As someone who wasn't bothered about Spotify, it made the band instantly known, as with your Release Radar thing. It's now on people's horizons in some way or another, which happened because I had approached it this way. 

There is no perfect beat. There is no finished thing. I had like 34 songs made for the album, and then I'd been like, all right, at some point instinct, these have this thing I like that I will decide on. That's what we're going to use. And then I'd said to the label, Oh, I've got these other eight tunes that I think are good, but I don't want to release.

And should we, seeing as though it's the pandemic, seen as though these, I think two of them got playlisted in a row because one, it could generate some money, two, it could generate some hype, and three, it would buy me time to work on this other stuff and finish it. And so that's why then we called it The Algorithm Smiles Upon You … 

LP: Because it did! 

Joe Mills: The mighty notion that one it did, but also that you can view it as a sinister way that a giant algorithm is watching over us, which I don't particularly feel comfortable about.

Also, the thing with my sister is this one use of Tinder. I also like Khruangbin, so we were playfully ripping because their first album is called The Universe Smiles Upon You. And so it was like, yeah, it'd be cool because if people get recommended this album to them, and they like it, and they're like, Oh, the algorithm did smile upon me.

And it's in the comments; the YouTube comments say it. And actually, we got our booking agent, this guy called André Marmot from Earth Agency. He got in touch and said, Yo, you came up on my Spotify, and it just seemed like destiny because of the title. All this sort of stuff, maybe I'm misquoting it. That then meant that my having made these 34 tunes wasn't wasted time.

And yeah, there was not this infinite potential. Still, there were a lot of songs that I saw no use in directly that ended up having a positive impact on us by just doing the process of it and not being really perfectionist or romantic about how good I wanted the end products to be. 

LP: The creation process that you articulated about how the tracks come together. How does that extend or apply to what you're doing in the live performance context? Because I think I understand, but I don't want to misunderstand. 

Joe Mills: In which sense? 

LP: When you're on stage, it sounds like you run a condensed version of that very same process. Like you're sampling stuff in real-time, and the artists, your collaborators, are reacting to that. 

Joe Mills: Yeah. Yeah. So there is a set list, and we're going to start on one song, and we're going to go into the second song. And then, we'll end up here, but the general parts between it are becoming less predictable and clear.

And the part of it that I operate in is that, for instance, with Middling, there's a singing part. I will trigger the singing part. So even if they're playing completely different stuff, that recognizable hook is there, which, for my hip-hop sentimentality, is something I want to maintain. If we're going to play Follow The Earworm.

I want you to hear the horn line and know this is that tune. I don't want it to be such an obscure remix of our own music that you can't tell. But what, for instance, happened when we played in London — we do this sort of a death drop of BPM. So Follow The Earworm is 134 BPM, and Follow The Earworm Part Two is 92 BPM, but they completely overdropped it.

And so they went from 134 down to 80 BPM, and it became this sort of like deep house, super slow, like the French horn had a big fat distortion. And we rocked that for four or five minutes just cause it sounded good. And we're talking on stage, and we maintained it. But then, when we came back from the tour, this was the first instance of me getting the stems from the desk and then being like, all right, I can take these parts, and then we re-released that.

So there's Follow The Earworm Part Three online, where we re-did the drums. But it just went off in its own direction. So, it still sounds familiar, but it's something new. And there's another element we'll venture into when we've got the actual technique down. We're sampling the drums on stage and granulating them, sampling the keys or the horn on stage and granulating them, and then allowing probably 10, 15, 20 minutes. Well, not that long, maybe 10 minutes to stretch out if we feel to let that go where it wants because I do like the idea of giving people what they want, but also provoking people. To make them wonder about what exactly it is they just ingested.

Do you know what I mean? Was that that song, and when did it change into this other song? And that's something we were playing on initially. We were only just getting started as a live act. So that thing in London was first, the first sold-out show we'd done, but also the first real resampling of ourselves and then putting that out as recorded material.

LP: It sounds like a recipe for spurring a passionate fan base because there's a lot of room for obsessiveness. 

Joe Mills: Yeah. I'm assuming you know who Makaya McCraven is. 

LP: Yes. 

Joe Mills: His sampled version of doing roughly the same thing is done in large chunks. Whereas mine is taking very, very tiny pieces and rearranging them, and his is like amazing loops. And when I saw him play live, the band was impersonating the pitching up and pitching down of a record between the four or five players, which is killer. But we're trying to do the opposite, using the computer to create the fluctuations on stage so that the musicians adapt.

Not that being pre-learned is a bad thing. The technical ability to do what they were doing is incredible. Also, because there was a bit on stage where all the crowd was shouting back at us, and I was like, yeah, I want to put that in the records, and I know he does that as well. And that, that, like you said, obsessiveness, that thing, I want people to be able to come and get something different from it, but also to be involved in some way.

These sorts of things I still need to explore fully. I also had this amazing idea where I would interview people before the show and have such a quick turnaround that I can program their interesting quotes and whatever and play it on stage so that you are fully part of it.

I do a radio show, and people sent me voice notes about positive ideas for the future, a world they want to live in. And then there's a real nice occurrence. We did an album launch in Berlin, and I didn't know one of my friends would come to the event night. And I'd sampled her talking about how she's Palestinian. She was talking about whether we are all going to fight over these little blocks of land we're born on. Or are we going to live in a better place? Are we going to take psychedelics in the garden and be drunk with ecstasy and this beautiful sort of poetic speech? And I played this on stage, and then she was there, and she was like, this was a singularly important, well, not important, but beautiful thing to hear.

I also want to manifest some way to get people integrated into it. It just gives us better ideas. If we're feeding back off the audience, It leads the musicians in a way; although it's still instrumental, it gives a narrative to the music in a way that doesn't have a rapper dictating it or a singer controlling it.

LP: I'm sure playing with people with a jazz background or a free improv background, I can completely understand how they can thrive in these moments and in this environment. But you also mentioned some people who played from a more classical background. I think I heard you say that. 

Joe Mills: Yeah.

LP: The knock on those musicians is they can't do it if it's not on the page. They never developed those chops. And I wonder if, putting aside whether that's true or not, or that's overbroad, but how is that experience different? Are you finding that it's a challenge for musicians of different backgrounds? And is there fertility in that challenge? 

Joe Mills: So Meravi, initially, in the pandemic, we'd done some recording sessions, and I went around to her house. And I mostly sat there in silence and occasionally hummed a few ideas to them. But Daron and Meravi did the whole thing. They just illustrated loads of dreams I'd had because I didn't realize the dynamic range of the French horn.

She explained to me that the attack is the thing that dictates the sound, and if you electrify it, you can't tell if it's a guitar, if it's a horn, or if it's a violin. And she could play as beautiful as a string line, or she could do a real grungy, punchin' sort of distorted foghorn. In my head, I was already like, she has to be in the band. I would love her to be in a band. But the initial thing that came up was that she couldn't improvise. But then also, neither can I. I am the least skillful of all the musicians. I'm also going against my principles. I don't like button-pushing. And I'm on stage pressing buttons that I've organized in a grid, in an alphabetical, Western alphabetical, left to right, top to bottom kind of thing, so I can play the parts as I go through. And then there's the granulated bits, which are separate. So when Meravi struggles with the improvisation bits, we can't get as loose. I'm also in that same field. So there's part of it where because we are operating roughly in. a range of chords or keys or whatever, but she's sat with me.

We're also next to each other on stage in this way. So we're like, all right, you try, and you move into this bit now. Daron is also standing next to her. So there is a, there's a nervy energy to it. Like when we venture into the stranger parts. But also, it's something that she wants to embrace.

It's also something that I walk into. So, that part of the official myth is true. She struggles more to improvise with it. But then she also turns up to all the sessions fully rehearsed. So she knows she's made herself and done this whole prep. The improvisers don't do as much.

They come up and like, yeah, you know, we're going to wing it. They're relearning their own improvisations. So they're struggling in a different way where she turns up fully in the mix. It's having to lean into that bit. Still, because the French horn is playing the catchy bits or the memorable lines when she goes off into this sort of sound effect department of the horn with lots of effects that she's also starting to learn, it complements them. It allows them to spread out more. And when we bring it back, it comes into the center where she and I are playing the more memorable parts. At the moment, it's working well. You'd have to ask her. She might say differently and that she struggles with loads of it or that she is embracing it. Yeah, they're completely different worlds. 

LP: It doesn't sound dissimilar from how song forms or, like the canon of standard works in jazz improvisation, right? There's the, there'd be the start of the song is the recognizable part of, say, an old Broadway show tune or something, and that's the launchpad for the improvisation.

But then it's there again at the end. So there, it's almost like there's a grounding effect that you all play for each other so that it doesn't become completely free. 

Joe Mills: Yeah. I don't want it to get totally free because I don't mind it like having flashes of it. And, when you contain it, and we only let out tiny wild 30-second bursts, it gets a much better reaction.

But Nir, the drummer, was previously in a free jazz band, and he kept saying that if there are more audience members than people on stage, it's a good show. All right, I'm not looking to scare the shit out of everyone and go so off piece that it's unlistenable to the general ear.

But this is a funny road we must walk now because I want old jazz heads to appreciate. The strange shit that we were trying to do, but I also want the trip park, hip hop, and even pop people to be intrigued, to hear what it is. We were talking to our booking agent about this the other day. Even with the bookings we want to get, do we want to play in small jazz clubs? 

We just played at a blues festival in Romania. I don't know what they expected, but when we were sound-checking, the engineer said, yeah, this gets pretty weird and psychedelic. I wasn't expecting that. And everyone else was a blues act. And so this is a creative decision we have to make now within where we play. I don't want it to get too stretched out.

LP: It's a very interesting point you bring up, and it's something that's come up, actually, two themes there that have come up a lot in other conversations I have here. The first is, broadly speaking, like the resurgence or the infusion of jazz into a lot of modern music. I'm a bit older than you are, but it's been something I was never expecting to see.

I grew up in an era where the conversation was, is jazz dead? Is there anywhere left for the form to go? And to see it cross with electronic music and with hip hop and to be a part of the cultural conversation again is so, like, it's just so gratifying. It's so healthy for the music and artists like what you're doing, a lot of artists across Europe in particular.

It also begs the question of where to present the music. Again, I've talked to many artists who have struggled with "where should we play? Do we play rock clubs? Do we play jazz rooms? Do we play nontraditional rooms and find spaces to present our work in a new way?" New forms are emerging and can be presented in new ways if the artist wants to go there.

Joe Mills: Yeah, so the booking agent who we're working with who sought us out, who I started to get to know a little bit, we only started working on last December. And he's writing a book, or he's just sent me the first draft of his book called Unapologetic Expression, which is about the resurgence of jazz, in London in particular.

He's from London, so you have to give his London-centric point of view on things, but there is a thought that basically because of Yussef Kamaal, the resurgence of that album, obviously championed by Gilles Peterson as well, but what that the attachment to grime and drum and bass and the recycling that through Afrobeat and jazz playing with Ezra Collective and so on, that this gave it like a massive sort of groundswell in London.

He talks about how you wouldn't have thought that a band led by a tuba player would sell out like a 5,000-person venue. In 2016, even when those records were the initial inklings were coming out, and then you got something like the Pharoah Sanders Floating Points album, which is a late-career classic for Sanders and a completely mad feat of, I don't know, creative writing in terms of Sam Shepherd because you get this, 38-minute ambient piece that's conducted for an orchestra, written by what a lot of people know as like a house producer. He got 'Best Album of the Year' in Time Magazine. 

And where you place this stuff, I don't know, but the bit that I, personally, from my sort of hip hop background, I've loved so far is being in a sort of small sweaty room with about 400, well, 300 people is the best that we've had or 250. And then it's tight playing in there because I don't know if ours will transfer too much bigger than that. Like we're playing jazz cafe in London. We're doing a UK tour in November and playing Jazz Cafe during Jazz Week. And that's going to be like 475 people. And I think. The general thinking is that to make money on the road, you need to be playing to more than 500 people every single night. No, I want to get beyond that, but I think that small, big fat sound system, big sub, tight room, preferably smokey, but maybe it won't be because it's the UK.

Like that, that kind of atmosphere is where I would want to hit it but where everyone else is playing. 

Floating Points is doing a one-off gig at the Hollywood Bowl with Shabaka Hutchings playing the Pharoah Sanders parts and with the Hollywood Symphony Orchestra. Yeah. So it's like, it's got some limits somewhere, but it's stretching out to places I don't think anyone thought it would go.

But from my Mancunian Northern English drum and bass upbringing, a lot of people like Photek referred to that as an extension of jazz. They were taking the Miles Davises, speeding up the drum breaks from the jazz records, and twisting it in such a way that although it's quantized, it was still like an exploration of the same veins.

And if you connect that to where everything was in 2016, I find it very interesting. People think it went away, but actually, it was just bubbling under. And like with the grime scene in the UK where people thought it had died in 2006. They were like, "Fuck it, if you guys don't want to pay for it, we're gonna keep doing it."

And that's the beautiful thing with the wider, new jazz communities, that people are getting credit for it. But as Malcolm Catto, who is the drummer in The Heliocentrics and produced tracks on the Yussef Kamaal album, said, "If the London jazz scene lives or dies, like if it dies tomorrow, I'll still be doing it."

And the real practitioners, the people who are actually about it, will continue to do so either way. Because although it may seem like I've gone from being a hip-hop producer to doing this, I was making this weird instrumental shit to no one, to no audience, in 2017. And if it goes away, and if the Spotify playlisting drifts off, it's like Nir is essentially my brother.

I'm still going to end up fucking around with this shit. Finding the patterns and the grooves and all that's deeply fascinating; it's much more interesting than a lot of other musical art forms. Well, for me anyway. It has such a wide variety and depth to it that it encompasses all of the things in a way.

LP: Endlessly malleable as well. Yeah. Yeah. Earlier, you mentioned DJ Shadow, and I wonder, in light of, and even especially the last five, 10 minutes of our conversation, how do you see yourself? As part of a lineage, or do you think of yourself as a collagist? What's the self-image of you?

Joe Mills: That's the exact term. I don't know if you can see it on the wall behind me; there is a giant photo montage. At some point when I was about 16, my dad was very much into David Hockney. David Hockney does these photo collages, and I did that before making music. I'm just making collage.

It's the easiest way to explain it. Although there is a musical element to it, the bits when it goes off piece and goes strange, which I also enjoy. And it goes a bit more hip hop, leading into my real expression. I just do collage.

 LP: You know why? 

Joe Mills: Yeah. The Shadow thing — I remember saying to someone, I think Endtroducing is the greatest work of collage in the 20th century. Why does it even matter? But I think it's because in an era of information overload or such an overwhelming amount of data or whatever, the colors and images operate in one thing.

It seems quite simple. If you instantly place this one thing on top of another, it just recontextualizes everything. And I just love that. I love the Warhol things, for one example. And the toilet, what is it? [Marcel Duchamp's] Fountain is another one. 

And I'm obsessed with this band called Company Flow as well. And they had that. El-P is now the main guy behind Run the Jewels or one of the half of Run the Jewels. But in their racks, it was always this cultural reference. Like I'll be like, Oh, what does this mean? What cartoon is that from? And where is it? This layering of everything. And so, from the early stage, I was doing that.

And I love it now — like it made the root of this conversation. Even the fact that you asked me if Lee Sedol heard of us? It's just something I took from a film, and then we were like, Oh, we should call the tune The Lonely Tears of Lee Sedol cause this ties into this. And it just, I just find it very enjoyable more than anything.

I have quite a psychedelically orientated mind. The synthesis or the creation of novelty by combining two things and how they give definition to one another is just an endless game of fun. Yeah, that's not the most articulate way of putting it.

I've sat down and played the piano, but it doesn't strike me as fun. I'd probably sample myself playing the piano. 

LP: It's fortunate, I suppose, that we didn't start with this part of the conversation because I think we would have spent the hour going down the rabbit hole of New York underground hip hop because as soon as you get into Company Flow and you start talking about the things you just brought up, I'm going to meet you there and we're going to talk about MF Doom and we're going to, we're going to, all that world, really what you're, a lot of what you're talking about is these artists, they build worlds.

El-P especially, his production, just every one of his produced tracks, sounds like its own little outpost in a universe of all his tracks. It is like he has a fictional universe he lives in. Maybe not as articulate as Doom's, but you can hear an El-P production. 

Joe Mills: Yeah, and also the way he would flip something, I find quite annoying because I own a lot of the records, and I'll be sifting through it, and I'd be like, oh, I didn't know I hear that.

Yeah, but that's the bit that's the collage. So, everyone is presented with this same world. And then he chose to make this dark, sinister version of it and/or interpretation of it. And I, yeah, I have this thing where I don't know how else to do anything now. I just found myself here wanting to do it. And yeah, it's just taken me to fun places. So, I'm not going to learn an instrument anytime soon. 

LP: One last thing before I let you go on the DJ Shadow point. I haven't listened in years, but just recently, like two or three days ago. I revisited Cut Chemist and DJ Shadow's Product Placement.

Joe Mills: Oh, mate, it's amazing. 

LP: I haven't listened to it, probably in 15 years. The crazy thing was I listened to it so much back in the day, so I knew every nook and cranny of it as it played. It was so fun to hear it again. And it's shocking, the level of artistry and skill. 

Joe Mills: That's one of if not the best DJ set I've ever seen is them doing that live.

LP: I never saw them do it. 

Joe Mills: Maybe it wasn't. They did parts of it live, but they got Bambaataa's record collection and did all the original breaks, but they included parts of that in it. And it was like the visual thing behind it. And I remember, so I'm friends with a guy called DJ Format, and I rang him.

And I was like, you've got to see this thing. It's he's more of a nerd than I am. He's a deep, like old school UK record collector, DJ, beat maker. And he, yeah, he went to see it the next night. It was like, yeah, you're right. It's insane.

But again, man, that, that world-building thing that my obsession with Shadow is the, and that like atmospheric recycling of essentially broken dreams into this beautiful one-off record, really. Like he never recreated it. I don't think he ever wanted to recreate it. That is such a unique skill. I'm fascinated by the same with RZA; how did he think the kung fu films and all the stabs and nonmusical things would go together with nine people ranting at each other?

How did he come up with that? And also that run, I'm obsessed with that run he did with 93 through 97. Yeah, just that, get your head down, like work as hard as possible while they catch the lightning in a bottle. Obviously, we're nowhere near the same worlds, but at the minute, this is our 'why.' 

By the time the third one comes out, we've recorded the fourth one, and by the time the fifth one comes out, the fourth one's going to be called Game Four. The fifth one's going to be called Jazz in the Age of Data. I want to get it to that point, and either the band will have taken off and will be flying, or it will have exploded, and there'll be five records, and that's it. 

LP: Well, I'm rooting for you.