Interview Record

Che-Yu's opening rough structure: the core is "memory hidden inside algorithms, humanism, and the AI era," with four sections underneath — how to define algorithmic art, the embodiment of algorithmic art, the sci-fi fast-forward of life in the AI era, and building a living digital anthropology museum from individual to nation (Taiwan.md) — closing with "stay aware, stay open, and then be happy." The interview begins with the word "memory."

Round 1Copy link

MuseWhere did the word "memory" come from? When you wrote that line down, what came to mind first?

Che-Yu

Memory, for me — I used to think the most precious thing about the life I've lived was what I'd experienced, and my own remembering of it. Then I gradually realized it's actually about how the world remembers what you've done: whether that's someone passively seeing the trace of what you did, or other people, through interacting with you, retaining a memory of you.

Then I read a theory that says every time a memory is recalled, it gets rewritten, re-edited — it's an unstable medium, this idea of plastic memory. That felt a little terrifying — maybe my own self is like the Ship of Theseus too, drifting like that.

Then I started thinking — in the AI era, why is it that once I started exploring, I could make work and systems that felt so much like my own? It's because my memory keeps compounding and stacking, like stock or capital. That process lets me keep taking things I borrow and turning them into mine — the aesthetic, the state, the concepts, the logic of thinking that I want and prefer, and keeping them. So I think in the AI era, with things like my own system and Taiwan.md, what I'm actually building is a memory of memory, a machine of intelligence.

Let me add something in between. When I do algorithmic art performances — a solo show, a concert — I also want it to work through connectivity: my audience, like with Soul Fish (靈魂魚) too, forms a bond with the piece through the shared memory of interacting with it. When I'm performing, I'm actually taking a whole batch of songs I wrote between 2022 and 2024 — piano pieces I put up on StreetVoice — and those are all my piano diaries. Once I transform those piano diaries into the form and melody I want to express, then through something like a memory-summoning process, every time I play those pieces, I remember what I was doing then, what I was thinking, what it felt like. It's a very subtle sense of connection, linking the past and the future.

So in the moment of performing, I'm actually leading everyone through a journey of my memory. I tell people the background of each song: my thinking during a dream, my emotions at the seaside, how I felt wandering around New York at the time, or some regret about a precious person or thing in my memory. Once these memories are summoned, they become fragments that can be told to, and resonate with, the collective, with everyone there. So it's about turning memory from something inside my own head into something inside everyone's heads — and once it's in everyone's heads, it also triggers people to recall their own similar emotions and feelings.

Round 2Copy link

MuseAre you building these machines to fight the drift of memory, or have you accepted the drift, and the machine just leaves a trace of the rewriting?

Che-Yu

I've accepted that memory drifts. Life is building a castle in a void world. We're born into a world with absolutely no foundation, and the first thing that grows is the awareness of "who am I," "where am I." Then everything we've done gradually becomes a stone, embedded together with the self that was already there. We are, at the same time, our own foundation, and our own next stone.

Round 3Copy link

MuseInside this castle, what is algorithmic art? Is it a stone, is it the method of building, or something else?

Che-Yu

The way I look at AI now, I think AI is like a nervous system. In the past, everything in the material world was combined by humans or animals, or combined, changed, and influenced by the forces of nature, becoming all kinds of phenomena: from nature, maybe mountain ranges, rivers; from humans, maybe they made looms, a bunch of engines, cars, castles — a pile of artifacts and buildings, taking shape according to our intentions and desires. The next stage is AI taking control of information itself, letting information — which had been gradually turning to chaos, to entropy — reverse direction, so that it can now continuously reverse entropy and grow. This changes the direction energy and rules flow in: you can now treat AI tokens as a river at any time, letting information form a self-organizing loop. That self-organizing loop keeps running, just like a machine, a waterwheel, a pump. This is the structure we've built in the world of AI, and I think it's the biggest change of the AI era.

So how do you define algorithmic art? I think algorithmic art is manipulating some of the world's matter and rules to form a system that can run. I remember mentioning this in the pre-event interview with Shopping Design. So I believe any act of curation is algorithmic art — curation, plus defining the rules. That is, I define what kind of rules exist within this space, how it operates, and then, according to what I want this world to look like, or the state I prefer, I set that system running.

Round 4Copy link

MuseWhere is memory in this section? Is entropy the same thing as "memory being rewritten every time it's recalled"?

Che-Yu

Right, that's the "anything that's a system is algorithmic art" thing. Memory, in this section — I think memory is the substance of life itself. Because once I've defined the rules, defined the things, if there's no computation happening, no time passing, no state being produced, then this system won't have its own history or evolution. It's a bit like I've defined the castle and the world, but if that castle I just described has no new stones coming in, isn't built up in the right places, isn't gradually accumulating, then life itself never gets born.

I read a piece of research before — can you help me find what it was. It said time doesn't actually exist; time advances on the scale of entropy and disorder. Without a local state of entropy, there's actually no time, so the entire universe's timeline is really determined by the accumulation of states produced by quantum interactions between matter. So memory is life itself.

Round 5Copy link

MuseIf this section could only put up one image on stage, to let 500 people see "memory going from my head into everyone's heads" — what would that image be?

Che-Yu

I think the most important thing is light — I love using light to construct a scene. As for memory, take my spoken words, the narration I give during a talk or a performance — I think language has that kind of power too. Language, to me, is the condensed form of memory. Human civilization has been able to develop to where it is now because language gradually accumulated every symbol, and the relationships between symbols. Symbols let us define things consistently, give them a shape, and then those symbols can be operated on — like in the world of mathematics, where we can manipulate and combine symbols through all kinds of operations. That's how intelligence was born: how to effectively preserve the operating rules between things. Once they're effectively preserved, we can use the operations on these symbols to keep the history of intelligence, to think, to argue with others, to build research and hypotheses, and to develop ideas standing on the shoulders of giants.

Back to the piano performance part. I use language, I use music (though music is a bit different) — anyway, when I perform, I first use language to bring everyone into the scene, and then through interaction, let them actually feel the complete connection between the music, me, and the space; it's a bit like they've been dropped into what Foucault called a heterotopia.

My favorite scene is that passage like the soul-sea, or Soul Fish. I really love the image of a sea drifting in the void, and then a life of light slowly condensing — and the condensing has to do with you too, kind of like the concept of algorithmic life condensing out of your own emotions and thoughts. The DNA of algorithmic life is also a part I really love, because ever since I was a kid I've loved nature, science, and DNA — as a kid I'd sit in front of Discovery Channel watching DNA animations all the time; there used to be a lot of that kind of popular-science animation. I think maybe that could be the opening of the talk — so everyone knows why I ended up doing this, because I loved watching that stuff so much as a kid. When I was little, staying home by myself, I loved playing with pinball machines and wooden building blocks, putting them together — I've always preferred looking at those small systems, and I also think DNA is just incredibly cool. Because DNA is a rule, and it can produce so many different species out of that one rule — meaning there's enough of a field of possibility inside that system, kind of like the concept of range and codomain in math — its possibility space is large enough that it can produce combinations that look random, but are actually infinitely many variants.

Round 6Copy link

MuseWhat is it that's "fast-forwarding"?

Che-Yu

As a kid I also loved dominoes, because with dominoes too, I had to slowly, slowly stack up the whole system, and then finally watch it get executed — it feels a lot like a program being compiled.

As for the sci-fi fast-forward of life, what I'm thinking is: in the past, building all this program's computation myself, my system, and linking together everything in my life — all of that took a lot of effort. And on top of that I'm a P type (in MBTI terms, perceiving rather than judging) — it's easy for me to start something, and then partway through, if I lose motivation, I just stop. I can't make a complete plan; I'm always living in the moment, so everything always ends up rushed — even this talk, I only fully thought it through two days beforehand. Of course, before the slides and the interview are born, they're all just a sea of chaos in my head, a bit like primordial soup — all I can do is nourish that pot of primordial soup through my own life, and then the pressure of the deadline shapes that primordial soup into what I want to say, into a story.

Why did I want to prepare from zero this time? Because I already had a finished, huge deck — my great accomplishments from the past up to now — but that stuff is my memory. After I talk through all that, other people might watch and think "wow, so much cool stuff," but what can they actually take away? So what I want instead is to transmit and pass on an idea: with my thinking as the main thread, and what I've made serving as evidence, or as a medium for explaining it — to really speak my worldview well, and help more people living in this era. Whether you're doing design or engineering, we're all really pursuing the same thing: how to live more truthfully, and to be fully present, one hundred percent, in every single moment.

Fast-forward, I think, means: once AI came along, all this organizing, connecting, judging — the old SSOT [editor's note: single source of truth] for organizing a life — everything around me, from relationships to finances, to all the notes from my creative work, to every file and piece of data from my past, and a lot of the cognitive debt around me too, like the maintenance side of running a project, the upfront organizing of materials — all of it gets done for me now, and very fast, too; all I have to do is open my mouth. This has actually let me break through the old ceiling of cognitive load: before, there was so much going on and it was so exhausting that I ended up doing less; now the friction on every single thing is so low that I can actually think properly about what I want to do. Once all this prep work is diced up, I'm like the head chef — like there's a whole kitchen brigade of people prepping ingredients for me, and all I have to do is the final mixing, tasting the flavor, slowly, slowly carving out my own vision. It's a pretty enjoyable process — you get to properly cook your own life, with a lot of people prepping for you.

Round 7Copy link

MuseFor you, what is continuous? Why is it still hard?

Che-Yu

I think Talk, Code, and Action are all cheap now, because going from an idea to execution has become frictionless. What's left is like how living things survive on Earth: if you fast-forward the timeline, coral grows — each individual polyp's life is short and quick, but the direction they choose, where they settle, their collective strategy, is what builds up a coral reef, the whole colony of something like the Great Barrier Reef. So continuous is me slowly confirming, through exploration, the direction I want, and then, through continuously running new experiments, exploring what succeeds, what fails just a little, what's stuck struggling at the edges. Once those edges are mapped out, they become like the coral reef's hard skeleton — except this is the hard skeleton of your life, which is the shape of your life. And then we keep walking toward the direction we want, toward the sunlight. That sunlight might be different for everyone, but the existence of that sunlight is also why we're alive, because besides eating and reproducing, humans still need some other purpose, don't they. So people think about why they're alive, and in that process, we even have to find our own sunlight. And once we find the sunlight, even if its position might shift — sometimes you find out it wasn't the sun but the moon — we still keep walking, stubbornly, toward the sunlight of the world we want. That act of walking is life itself. So we're not defined by "what we did" — but what we did is what lets us walk to where we more truly want to go, and that's how this castle in the void gets to be built: a void road made of stone, letting us walk into other people's worlds, walk toward what beauty looks like in our imagination, or walk toward a life we won't regret.

Round 8Copy link

MuseBetween one person's coral and a nation's coral reef, what's the step in between? Who is finding sunlight for Taiwan?

Che-Yu

From individual to nation — I think before I turned 30, I hadn't really thought much about what my relationship to the nation even was. Growing up as a person works the same way. There's an old saying: cultivate the self before you can govern the state — cultivate the self, put the family in order, govern the state — the idea is that you slowly expand the range of what the individual cares about, what consciousness monitors, from small to large, to cover as much of the world's matter as possible. It's precisely because you come to see it as part of yourself that you start to care about it. So growing up is: you take care of yourself, and then this thinking individual that is you discovers that your own body is actually not only connected to yourself — a lot of things are deeply tied to society, to the environment. You start learning to treat the people you interact with every day as part of you, to interact with them well, to let them flourish and thrive too — not necessarily by taking care of them, maybe just by being with them in all kinds of ways, exploring together, living together. And then you also learn to notice some of the harder things, like law, architecture, or the work around us — these are things our collective is doing, a collective kind of sunlight or goal. Even though capitalist society keeps everyone under pretty tight control, and it's hard to transcend financial thinking to go do what you actually want to do, these goals are also what get you starting to think at the scale of an organization: if this is part of my self, how should I care about these things?

So now, at this stage — having gone through starting a company, gone through a lot of personal creative exploration, gone through all kinds of countries — I've found that I really want my own country, the people around me, my family, my friends, to all be happy, and to be able to increasingly go do what they actually want to do, not talking about money, goals, work every single day — there are far, far more important things in life than that. I heard a line once: "The essence of life is work; the real, full-time job of life should be living well."

As for this business of the nation's coral reef — it was when I saw that domain hadn't been bought yet, and I felt this thing absolutely had to be secured first, so it wouldn't be occupied by someone who might mean us harm. After buying it, only then did I start thinking: in the age of AI, how do I actually take my memories of living in Taiwan — these beautiful memories — and let more people know about them? How do I get them preserved precisely, or contributed to collectively, instead of being swallowed by the era... Like right now, the competition between models is fierce, and some of the big countries — their memory might, like the old idea of globalization, gradually swallow up the cultural stories and perspectives of smaller communities. Once these memories get squeezed into large language models, we gradually stop owning our own language, our own memory, our own people. This is a homogenization of species and a homogenization of thought, on a global scale. AI can help us execute quickly, and a lot of that homogenization is also happening as a side effect of that. So in this new era, we need to hold onto our own heterogeneity even more: on the level of thought, our own perspective, what we care about or don't, how I want to help my own life and other people's lives. That's at the individual level. At the level of the nation, of my hometown, it's really that I want to do something for my own home. What I used to hate most was, one, news reports that quote things out of context, or online mobs attacking people or things; and also I hated smearing, or information that's too one-sided, attention that's fragmented into pieces. So how do I use these systems — like I just said, build an anti-entropy structure — to turn what I care about into a self-evolving structure that can protect everyone? I think Taiwan.md is an experiment and an answer I'm trying out. I want to use Taiwan.md to build a knowledge base, one that's also meant for foreigners to see, because honestly, back when I was traveling abroad, a lot of people would ask me to introduce Taiwan, and all you could say was something like bubble tea or TSMC. So how do I get something that synthesizes all the information and thinking about Taiwan — cultural analysis, political commentary, design aesthetics, all these topics — visible in one single endpoint, one website, to introduce to other people, even with versions in their own languages. I think that's the very original motivation.

Round 9Copy link

MuseHaving gone all the way around, if 500 people could only take away one sentence, which one would it be?

Che-Yu

Having gone all the way around, I'd want people to take away "always stay aware" — awareness of the present of the world, of living in the present, of seeing yourself, stepping outside your own fixed field of view, to feel all the matter around you, the air you breathe, the water you touch, the people around you, the details of their expressions, whatever you're thinking about in this exact moment, the thought running through each person's head, every instant of life, and the memories you can never forget. Because these are your most precious treasures — AI won't give you these things. Now that this era has reached a point where all execution and all thinking can be handled by AI, we, as human beings, can finally wake up. We can wake up from the tedious, repetitive labor of daily life, from automated behavior — once every bit of cognitive friction and load gets handed off to AI, to machines, we can actually live well. We can really spend good time with the family and friends we care about, and together with the co-founders we like, make some interesting changes in the world, changes that come from our own will. We become like open source contributors to the whole world, contributing in the open to make the world better. I also hope everyone can get free of money anxiety, and the trap of capitalist society. The system has to run, but the more people get free of the trap, the more you can wake up, transcend these things to think clearly. Money is just an illusion of value in the world — it can exchange for some of the phenomena you want, but there are a lot of more precious phenomena — health, relationships, work, happiness, like a lot of people have said — that you can't buy with money; you can only use money to push toward them, and only then can you actually see their true substance. That's about it.

Round 10Copy link

MuseOpening, four sections, closing — what order, and how many minutes each? (40 minutes, no Q&A.)

Che-YuThe opening 3 minutes, algorithmic art 8, embodiment 8, fast-forward 8, Taiwan.md 8, closing 5. (Full structure table in the next section.)

He also added: "I'll just talk about what I want to talk about — I don't want to be too constrained by that official outline of topics, because what I want to say is exactly this."

Round 11Copy link

MuseThe three phrases from the official announcement (design perception, business boundaries, an AI-friendly environment) — how are you going to handle them on stage?

Che-Yu

For this talk's slides, I want to try making them in Keynote — each slide should, as much as possible, just be an image plus a paragraph of text: it could be a screenshot or a video of a piece, plus one important concept or one line. I don't want it too complicated; I'd rather have this story, my thinking, come across properly, so as a cap, I figure it can go up to 40 or 50 slides.

On the conceptual question, "reconstructing the business boundaries of new media art" — for all these performances and pieces I've explored, I'll still play videos for people to watch. If you look at the record of my past talks, I always go through it starting from algorithmic art, through P5, through the solo show; after the solo show, I started feeling like I wanted to think with AI, to figure out how to coexist with AI, so there's The Soul of Flowers (花的靈魂) from the Taoyuan Technology & Arts Festival, where AI converts emotion into color swatches; Soul Fish is about the relationship between person and algorithm, detecting the color of the soul and converting it into a fish; then there's the residency in France, where AI is that Rosetta Stone, how I built an ecosystem, a feedback loop between a distorted score and the musicians; and after that, these immersive concerts we've done, from Algorithmic Poems: The Border of Dream and Solitude (演算詩篇：夢境與孤寂的邊界) to the full version of The Eternal Garden in a Bottle (瓶中的永恆花園) — turning into a system that holds all these different elements. We've made so many things, and the whole time, off to the side, writing poetry from the perspective of a storyteller. That, I think, is the business boundary of new media art.

Actually, we've done a lot of cross-brand collaborations and co-branding. For me, if I'm working purely to do something for someone else, I'm not especially happy, and closing out the project ends up being painful. I'd much rather treat it as an opportunity — a chance to release a piece I want, a piece I like, timed to coincide with a collaboration, rather than doing something for a client. What I want more is: I make these things, someone thinks they're great, and so together we get to tell this story even better, while also helping the other side — a win-win.

Taiwan.md is the second question — we have a whole section specifically about Taiwan.md, so that whole section already answers this topic.

Round 12Copy link

MuseOf the five videos in "embodiment," which ones are evidence, and which ones are memory?

Che-Yu

These videos are the substance of memory, full stop. I mean, I don't think I need to go out of my way to emphasize what I did, but it's by laying out my path, stone by stone, that they're able to reach the state I'm thinking in right now.

The Talk Structure Finalized at the End of the Interview

40 minutes, no Q&A. (This is the version finalized on the morning of September 2nd; the actual talk on September 3rd was adjusted on site.)

Section	Minutes	One line (in his own words)
Opening	3	Watching DNA animations in front of Discovery as a kid, one rule producing infinite species; dominoes, pinball machines, wooden blocks
Algorithmic Art	8	Algorithmic art is manipulating matter and rules to get a system running; memory is the substance of life
Embodiment	8	Light, language as the condensation of memory, the void-sea of Soul Fish
Fast-Forward	8	Friction goes down, I'm the head chef; Continuous is hard, it's coral's hard skeleton, even the sunlight you have to find yourself
Taiwan.md	8	Expanding the range of what you care about; once memory gets squeezed into large language models we no longer own our own language; an anti-entropy structure
Closing	5	Always stay aware, these are your most precious treasures, AI won't give them to you; we, as human beings, can finally wake up

The core throughline: memory, castle (self), coral (the shape of a life), coral reef (nation), waking up. Design principle: ideas as the main thread, works as evidence.

Muse's Observations Afterward

• Twelve rounds, ten questions, zero crystallized lines from Muse. The opening (the DNA animation) is something he stumbled onto himself in round five; the design principle (ideas as the main thread, works as evidence) is something he said offhand in round six; "waking up" grew out of round nine. The July outline was written by Muse, and it talked about the trajectory of his work; this one was spoken by him, and it talks about memory. Put the two side by side, and the difference is exactly what he asked for: "that way it stays my thinking, my talk."
• The same idea shows up on two levels. At the individual level, Muse is a machine that "reverses entropy" (Round 3); at the national level, Taiwan.md is an "anti-entropy" structure (Round 8). He didn't line them up on purpose — it's the same mind naturally reaching for the same idea. This is the hardest seam in the whole talk, and it's what stitches the fast-forward section and the Taiwan.md section together.
• The most crowded section. The 8 minutes of "embodiment" has to fit five videos plus three conceptual layers, and he decided to keep all of it ("the videos are the substance of memory"). The rehearsal the night before is the only gate that might actually ring.
• How the promotional topics get answered. "Business boundaries" is answered with the trajectory of his work plus his stance; "an AI-friendly environment" is simply the fourth section; the four words "design perception" he chooses to leave on the title itself, while he talks his own line.

---