Knowlege Structures

 

 

Ryan Ahn (00:00)
Hello, this is the Special Team Matchup Podcast. I'm Ryan On and this is my co -host Zongjun Li

Zongjun Li (00:04)
Hello

everyone!

Ryan Ahn (00:06)
and we're the co -founders of Sanko Macha products.

Zongjun Li (00:09)
we launched this podcast to discuss our learning journey in Matcha, share some startup stories and interview experts. So today we are going to talk about knowledge structure. Ryan, since you come up with the topic, can you elaborate a little bit more of what you mean by knowledge structure?

Ryan Ahn (00:28)
Sure. So, knowledge structures are how we view the world and filter all the new information that we pick up every day. So like Copernicus discovering that the sun is at the center of the universe was a new form of knowledge structure. And like when you would acquire new information about the cosmos, it was a way to reframe and rethink about it.

So the way I like to think about knowledge structures and matcha is if you were having a conversation with a three -year -old about matcha and you wanted to talk about things you were interested in, things like cultivars or terroirs or the history of it, they probably wouldn't understand what you were saying because they haven't built that scaffolding, the knowledge structures yet to interpret that information and acquire the knowledge to fill that scaffolding to then appreciate what you're saying.

Different people have different sets of knowledge structures. Most people within certain communities have overlapping knowledge structures, so they view the world in a similar way. And sometimes there's conflicting ones, which is where you see a lot of debate. It can sometimes lead to revised knowledge structures or even new knowledge structures.

Zongjun Li (01:39)
Yeah,

so the universe example is a pretty interesting one. Basically, what you mean by that is that it's certain things that people view as ground truth. And with that as a foundation,

a conversation can happen in a logical way. Otherwise, you'll be just talking over each other and it's meaningless.

Ryan Ahn (02:15)
Yeah, exactly. And to make it super tangible, let's say you're a wine expert and have a lot of knowledge about what makes certain grapes that are grown in certain vineyards very high quality. And let's say we're talking about matcha and I say, yeah, there's no irrigation on the plantation. Someone who has that overlapping knowledge structure about wine and irrigation will know that that probably means you're going to have higher quality grapes or in this case, higher quality tea.

because they have those cause and effect relationships in their mind. their roots are going to grow deeper to access more water, and then it's going to have access to more nutrients, and then it's going to result in a higher quality product,

Zongjun Li (02:57)
but it doesn't necessarily would happen in that way, right? For example, why people with different knowledge structure, for example, a biodynamic, they will argue that it's the correct moon cycle that provide the best yield. So, different language structure, even though you see that it makes sense with their...

Ryan Ahn (03:09)
Hmm.

Yes.

Zongjun Li (03:21)
own community doesn't necessarily translate to the other.

Ryan Ahn (03:25)
Yeah. Oh, exactly. That's a fantastic point. I was visiting a biodynamic winery in Bordeaux last summer. And, you know, to me, it always meant biodiverse. That's what I always thought that that meant. And biodynamicism was way more philosophical than I ever realized. Because this lady started talking about water memory and spraying chemomile over the plants, but like only on the Tuesday. It's like, ooh, wow. Okay.

Zongjun Li (03:37)
No, no.

Rude.

Yep.

Ryan Ahn (03:54)
We have very different underlying assumptions about cause and effect relationships here.

Zongjun Li (03:58)
Yeah, well, it's like the, you know, pre -discovery era of chemistry and elements, right? Like people think everything is made up by, you know, wind and fire and gas and so on and so forth. And, you know, it's like one other example will be pre -modern medicine. And, you know, for example, in China, traditional Chinese medicine.

they viewed the world and the cause -effect of sickness and treatment very differently than modern techniques.

Ryan Ahn (04:35)
I feel like a lot of the conflict is this push and pull between conflicting knowledge structures. So to bring it into the article about knowledge structures for learning about matcha and growing your interest within matcha, traditionally, it's all been very top -down style knowledge structure imposition.

Zongjun Li (04:41)
That's right.

Ryan Ahn (04:56)
into a beginner's mind. When you study Japanese tea ceremony, you learn things in a very specific order. It's always the same order for all students. You're not even allowed to talk about your studies or teach them to other people until you've been licensed and given permission to officially do that. And it was a very linear path. And what you see today outside of tea ceremony is the exact inversion of that. It's bottom -up.

organic. It's what's fed to us on our Instagram posts and Facebook and TikTok. And it's these little like little sound bites, little digestible bits of information that then we have to structure ourselves versus this very top -down approach. So, Zongjun what are some of the merits and demerits of each approach? And what has your experience been learning things in both ways?

Zongjun Li (05:50)
Yeah,

so I follow a pretty traditional path in learning tea, both Japanese tea and Chinese tea, in which I follow a instructor, a teacher, a laoshi, who will come up with an agenda or a syllabus.

the topics, subjects, and knowledge that will be teached in each session. And throughout the way you learned the subject from a very linear way. So like starting from the very basics and then work your way up to more complex topics. It's very long and tedious frequently.

for a lot of the things that especially takes a lot of practice and memorization, recitation, muscle memory. But you end up having a very comprehensive view on the thing as a whole, right? Like you not only be able to learn about how to brew tea, especially in Chinese tea ceremony, but you also learn about how to grow tea, what kind of tea.

that you are drinking and how they are made and what's the origin, what's the history behind it. In Japanese tea, you learn less about the tea itself, but also a lot of things about calligraphy, flower arrangement, and Japanese aesthetic and philosophy behind the tea ceremony. So it's very comprehensive and

you essentially learning a shared vocabulary that you can make reasonable and logical conversation with other people learning the same subject. Versus in modern days, when you Google random subject on the internet or some random topics about tea on the internet, you probably end up running into 10 essays, articles or podcasts.

that doesn't necessarily use the same language and oftentimes making contradicting argument. So in a sense, your viewpoint gets highly diversified, but on the other hand, if you do not have any or enough formative experience about the thing that you're trying to research on, you get confused.

Like you don't know what is true and what is not true. And sometimes like even with citations in the articles, like frequent articles being cited that was written back in Qing dynasty or Ming dynasty, what I end up learning is that you really need to go back to that original book and read that phrase. Like people oftentimes twisted the terms or...

Ryan Ahn (08:28)
Mm.

Thank you.

Zongjun Li (08:52)
or use their own terms or sometimes they chop a sentence off from a paragraph for their own usage. And it's very, very confusing. It's very confusing.

Ryan Ahn (09:05)
Yeah, for sure. I mean, that's been my experience too. I also had a very linear path with like a set curriculum at the Penn State Tea Institute. And then also as a student of Japanese tea ceremony in Omotesenkei. And you know, I still get confused, right? You still come across contradicting information. And we grew up during a very interesting era in tea as it was coming onto the internet.

I mean, I remember the days when it was really, it was almost impossible to learn anything. There were like two people talking about matcha on YouTube. I it was like Teavana and someone named like Dr. Tea or something. It was really hard to learn, right? And I've actually gone back and watched some of those videos and I'm like, oh my God, wow. Like some of these people, not Teavana, but some other people, like were really just spewing misinformation that's been like,

Zongjun Li (09:43)
That's a great name.

Ryan Ahn (10:00)
scientifically disproven and some of it's harmful. Like one of it was, there was one guy saying the first brew of a tea extracts all of the caffeine and then in the second, third, fourth subsequent brews, there's no more caffeine. That's actually just flat out false. There's some really interesting research done by the Camellia sinensis guys up in Montreal showing caffeine extraction well into the fifth brew and it, you know, takes

Zongjun Li (10:15)
Yeah, a

Ryan Ahn (10:29)
on the order of 10 minutes to remove most of the caffeine in a tea. So it's demonstrably false.

Zongjun Li (10:43)
end up hiding in these false articles. Sometimes some information that takes years of research or years of learning can be just in a discussion in a random forum between two people. really takes a lot of time and experience, but...

The beauty of internet is that it's a very flat learning trajectory. You can learn a lot of things with very little effort. Or a lot of effort. I mean, a lot of effort in researching and verification, but very little effort in, you know, I don't know, like practicing or Zen learning or observation.

Ryan Ahn (11:17)
Yeah.

you

Zongjun Li (11:31)
or sometimes for someone that takes a good fortune to experience a certain type of tea, but who ends up writing a very comprehensive flavor profile review of the tea and also all the back knowledge about the tea. You don't necessarily get to experience that before the internet, but now you can just read them for free.

Ryan Ahn (11:59)
Yeah, I actually find it, it's really interesting because a lot of the romance of drinking matcha, studying matcha, learning about it, a lot of the marketing around the romance of it is this ancient practice, these ancient rituals that have been ceremonial practice for hundreds or thousands of years. if you want to participate in that practice,

It takes years to learn Japanese tea ceremony. You don't really learn a lot about tea. You know, it's kind of like deciding you want to become a sushi master. Like there is a very preset path that you have to go through and a lot of gatekeepers. And, you know, that's how matcha is marketed in the front. But in the back, the way that most people learn about it, it's very surface level on the internet with these different...

Zongjun Li (12:38)
. . .

Ryan Ahn (12:48)
you know, sound bites and, various blogs and sifting through this information, with that ever having gone through that experience. And I find that contradiction interesting.

Zongjun Li (12:55)
Yeah.

Yeah. And also the perspective and the specialty of the instructor really matters. Like you don't end up learning about how to grow tea or farm tea or what's the difference between one geography and another over that side of the mountain in a traditional tea ceremony school. They don't necessarily teach you that.

Ryan Ahn (13:14)
you

Zongjun Li (13:21)
But it's really helpful to understand that, to be able to discern types of tea that you like and make decisions on drinking tea or experiencing different flavor profile. And understand the underlying mechanism of how certain flavor can be presented in some teas versus another.

Ryan Ahn (13:46)
Yeah, I mean, matcha can exist in the world entirely devoid of ceremony. And I think that's what more and more vendors are doing now. And I know that's personally where our interest lies as well, where it's a specialty product, not necessarily a ceremonial product.

Zongjun Li (14:06)
Yeah, it's more of a ceremonial element. It's not even like a product of the ceremony. It's, yeah.

Ryan Ahn (14:14)
True, true, true, true. Well, the

finished product is.

Zongjun Li (14:17)
Yeah, well, supposedly the finished product is the experience as a whole.

Ryan Ahn (14:23)
Zongjun, in your learning journey, what knowledge structures have you found particularly

Zongjun Li (14:28)
I practice compound and recurve archery and it's not

only about how to aim the arrow in the target and make sure the arrow hit the target every time. It's also about the signs of the bow, the signs of the movement of the arrow, and also the history of different types of bow and how they end up converging or diverging throughout time in different cultures.

A lot of things, they share very similar kind of structure, even though they're totally different categories. And I find that super helpful in trying to have a deeper understanding of the world in general.

Ryan Ahn (15:22)
Hmm, that makes sense.

Zongjun Li (15:24)
What about you, Ryan?

Ryan Ahn (15:26)
Yeah. there's a lot of knowledge structures that you can borrow from various disciplines. So a lot of the stuff in specialty coffee or in fine wine really apply to tea. A lot of people talk about the specialty coffee movement as where coffee finally elevated to the status of a fine wine. And they used a very similar set of terms and vocabularies such as terroir, talking about cultivars, talking about...

growing methodologies, blending methodologies, the concept of vintage and how that can play an impact on the tea. So I find those very useful, especially around the flavor or sensory experience of matcha. And then also a lot of biology and chemistry as well can really help you understand matcha much better, almost from a food science perspective.

makes it unique, what makes it different, how you can brew it better, get better flavor expression, and understanding quality in this way. And then you can even zoom out to more economic factors, the globalization, the industrial revolution, and how that changed tea and tea processing. It's actually like where I find the greatest pleasure in learning about matcha is the fact that

that you can layer on almost endless knowledge structures and still find something interesting to learn.

Zongjun Li (16:49)
Yeah, so I guess, you know, with modern technology, you really end up demystifying a lot things that in the past being viewed as mystery or requires, certain kind of...

either personality or kind of learning trajectory or a very fixed community to be able to master. For example, like when it comes with the game of Go or Weiqi, it actually has a very similar kind of school structure to channel you. You have different schools of

Go in China, in South Korea, in Japan, and they all take very slightly different approach, but overall very Zen learning kind of approach, observation learning to teach a student how to learn the art. But with artificial intelligence, that really got essentially destroyed, right?

What once was a type of art that cannot be computationally solved is now, you have solutions to every kind of scenarios and machine is so much better than human.

so, Ryan, what do you think, you know, how do you assess the creation of all the impact of AI in, you know, all of these traditional art? And do you think it's generally helpful, positive? Or do you think that it really kind of deconstruct what is being viewed as a form of art?

Ryan Ahn (18:50)
So just to give a little bit of context, both Zongjun and I have professionally worked in AI and flavor development and new product development where we were optimizing flavor profiles of food and beverage products to better meet the consumer preference. So constantly discovering new flavor profiles and then working with formulators and bringing those to life.

being in that world and seeing it, the AI has superhuman creativity. AI in general are able to test more outcomes and more questions than humans could ever even think of. And it's changing the knowledge structures that we have today. AI has fundamentally changed strategy and knowledge structures within the game of Go. It was able to come up with moves that no one had ever thought of.

thought of before, they were completely unintuitive, and it was not obvious that they were a good idea until partially into the game. And we're starting to see this in the food and beverage industry, where you start to have superhuman creativity. And there's no reason why that couldn't be done in matcha. There's no reason why it couldn't be done in things like blending. There's no reason why it couldn't be done in the way we storytell.

and coming up with new connections, new things to pair matcha with that would make it delicious, new ways of improving quality, even down at the farm level. So that world is completely open. I would say that the existing LLMs are decent. I mean, we use LLMs for this podcast and in our content.

to refine them. Everything is written by humans, but then we'll have the AI challenge the ideas, ask new questions, broaden the scope of our own thinking. And in society in general, there's no reason why we shouldn't use these tools to expand our thinking.

Zongjun Li (20:55)
Ah, all bow to the AI Zen Master. Yeah, it's really true. With the technology we have nowadays, a lot of the things that, a lot of the man -made barriers in the past and now just breaking down and access to knowledge and access to research has never been so easy.

Ryan Ahn (21:17)
The

unfortunate side effect of all this generative AI in the short term is probably going to be a lot of junk. It's never been easier to write an article, put up a website, create a blog, and have ChatGPT write the article and then have some AI -generated images on top. And you're actually already starting to see that.

in the matcha world. So it's going to create more junk for people to sort through, which makes the importance of knowledge structures even better. Because the scale of tomorrow that's going to create experts in specialty matcha is going to be those people that are able to discern good and bad information on this learning journey.

Zongjun Li (21:49)
Yeah,

information entropy. It's terrible.

Ryan Ahn (22:09)
Yeah.

And it's a shame because it's commercially motivated. And that's the beauty of learning something in the traditional sense, where it's very top down. Like if you enroll yourself as a student in Omotesenke or Urasenke there is no commercial interest mixed in that education. Whereas people putting out content, I mean, including ourselves, we have some commercial interest, whether it be direct or indirect.

in influencing outcomes and the knowledge that people are learning.

Zongjun Li (22:42)
Yeah, absolutely. Nowadays, if you research some topics about tea, especially in Chinese, you probably need to flip at least two to three pages with all sorts of Taobao links in trying to sell your tea pots. And eventually you will be landing in some useful pages. It's very painful.

Ryan Ahn (22:56)
Thank you.

And just for full transparency, the reason we're doing this podcast is we see a huge potential for matcha as a specialty product outside of tea ceremony. And we think a rising tide will lift all boats. We'd like to create products to mill matcha fresh or to simplify matcha production or maybe even start our own matcha brand. But within the context of this podcast, we want to keep it totally non -commercial. Of course, you can't...

Zongjun Li (23:31)
Yeah.

Ryan Ahn (23:33)
eliminate all bias, but being very transparent in the tea world is a practice that should be more common.

Zongjun Li (23:39)
Yeah, we haven't inserted our podcast link into any Taobao or Amazon shop yet.

Ryan Ahn (23:45)
Yeah. No. So I guess to conclude this, Zongjun, the last section in this blog post was around creating new knowledge structures with research. So outside of things like AI, where do you find some interesting... Where's the bleeding edge of matcha knowledge and new frameworks that we can use to approach learning more about...

Zongjun Li (24:08)
Um,

that's a good question. Bleeding edge

So yeah, so I guess, you know, much much a largely nowadays is it's still in the very pre specialty state. A lot of the things that we are we essentially care about, right? Like terroirs and cultivar difference and you know, like.

growing techniques, even though some of the information are already available, there is a lack of categorization or systemization of all of these knowledge. There is no...

there is no coherent vocabulary in communicating your preference or your assessment of certainty. So I think that needs to happen sooner rather than later, especially if we want to really start to elevate the status of matcha into a specialty product, meaning that you can't...

have transparent labels in different matcha kind. You can educate your customer in a sense that they will be able to know exactly what kind of tea do they like. And if it's this kind of tea, what's the production day, what's the manufacturer region, what's the maker, all of those information are still largely...

unavailable in a lot of teas.

Ryan Ahn (25:59)
Yeah, for sure. Other areas that I've noticed too that are on the bleeding edge is also around nutrition, particularly around degradation and the importance of matcha freshness. And running studies looking at the effects of oxidation, how quickly does oxidation happen? How quickly is there a flavor change? Would a consumer notice? How quickly do the nutrients degrade?

There's some really interesting studies now, but it would be very interesting to see more in the future. As well as around like cultivar differences, are there caffeine differences? There's not really well documented flavor differences between them, as well as nutritional differences and caffeine differences as well. We'd be very interested.

Zongjun Li (26:44)
And yeah, and

what's the cause of those differences?

Ryan Ahn (26:48)
Yeah, for sure. What's the effect? Why would I...

Zongjun Li (26:52)
Yeah, I guess a lot

of these things that required new discovery won't necessarily be replaced by technology like AI. Yeah. So I probably would want to argue that with artificial intelligence, it doesn't necessarily create new knowledge. It doesn't necessarily...

revolutionize an existing knowledge structure. It will help people in doing so, but in order to construct a new knowledge structure or revise an existing one, it still requires human intervention.

Ryan Ahn (27:37)
Yeah, and new research, just flat out new research. There's a lot of low hanging fruit out there to make matcha radically better. There's a very interesting company in the specialty coffee world called Third Wave Water. And these guys just did tons of empirical tests. They took some theories that they knew about extraction, ran a bunch of tests with different water chemistries, playing around with the amount of magnesium and calcium and chloride.

Zongjun Li (27:39)
and new research.

Ryan Ahn (28:04)
in the water and optimized it for different roast profiles. They have one that's specifically optimized for espresso. You mix these mineral packets into water. Water chemistry has a huge impact on the way tea extraction happens. It feels like almost no one's talking about how to optimize water chemistry for matcha. Because other than matcha itself, the only other real ingredient that you have any control over is the water.

and it seems very low -hanging fruit that isn't getting enough light.

Zongjun Li (28:36)
Yeah, that's very interesting. Let's go.

Ryan Ahn (28:39)
So third wave water, if you're listening, we would love to work with you on setting up some tests to optimize some macha.

Alright, I think that wraps it up for this episode. Thanks for listening. Please follow or subscribe wherever you listen to podcasts. Please give us a five star review and we'll see you on the next episode. Thanks.

Zongjun Li (28:56)
All

right, thank you. Bye bye.

Ryan Ahn (29:02)
Bye.