content lead @KaiaChain | building the stu·dio (스튜·디오) | daily news & commentary | seoul 🇰🇷 x 🇿🇦
Seoul, South Korea
Joined November 2022
- Tweets3.6K
- Following1.1K
- Followers842
- Likes14K
Pinned Tweet
This one is a special one for me. Being given the opportunity to co-author a report with @antoniojmadeira and the @redstone_defi team is a huge honour, but moreso the thought and effort that went into crafting this report is what I'm proud of.
With the ever-growing volume of crypto research, it's hard to separate the wheat from the chaff, and I firmly believe that the strongest case for producing compelling media is eschewing what everyone else is doing in favour of something more abstract and tactile.
From the brand-agnostic risograph aesthetic that pays homage to our Korea-focused content using Taegeuk red and blue, to the subtle Hanji texture when you feel the flyer in hand, I intentionally crafted this report to feel as good in hand as it does digitally, and I hope that everyone who reads it feels that.
Meanwhile, Korea's DABA legislation remains stalled and speculation abounds over who gets to issue the won stablecoin. This report aims to answer some of those questions, providing key data points and context for institutions and allocators monitoring KRW stablecoin developments, while also forecasting potential scenarios when, rather than if, DABA is passed and the won stablecoin is introduced.
The Seoul Standard. Enjoy.
The rest of Asia has moved. Korea is still deciding.
Hong Kong licensed its first stablecoin issuers in April. Japan's yen stablecoin has been live since October 2025. Singapore is writing its framework into law. Korea passed its tokenized-securities law in January and reopened its exchanges to corporations, but the Digital Asset Basic Act, the statute that would license a won stablecoin, remains in committee over a single question: who gets to issue it.
While that question sits unanswered, net stablecoin outflows from Korean exchanges have run for eighteen consecutive months, about $10.4 billion in total, almost all of it in dollars. The demand for onchain assets is there. It is being served offshore.
Today we are publishing The Seoul Standard, a joint research report with @redstone_defi written for the allocators, funds and institutions weighing Korea. It covers:
- A decade of Korean digital asset legislation, from the 2017 ICO ban to VAUPA, and why each crisis produced the next statute
- What DABA and the enacted STO framework require of institutions, and the bank-versus-fintech dispute stalling one of them
- How a won stablecoin, tokenized deposits and tokenized securities could put inert KRW balances to work onchain
- Bear, base and bull scenarios with dated triggers, plus a sourced risk register
Read and download the full report (KR/EN): blog.kaia.io/the-seoul-stand…
everyone in korea is talking about pickle (피클) right now
the story is this:
it's six people in a house in hillsborough vs korean cyber giants like naver, kakao and skt
요즘 타임라인이 피클 초대장으로 도배됐네요
힐스버러 집 한 채에서 6명이 네이버, 카카오, SKT랑 붙는 중
1
2
135
we edit everything agentically at @KaiaChain and @RatioFX now:
design guides with built-in spec sheets, running @HyperFrames_ on opus 5.5.
why we made the switch:
1. brand consistency by default. fonts, colors, caption styles and safe zones live in the spec sheet, so every cut comes out on-brand without anyone policing it.
2. speed. a talking-head clip goes from raw footage to captioned, branded edit in a matter of minutes, so we can capitalize on momentum.
3. two(+) brands, one small team. the same pipeline runs on both orgs with different guides plugged in, so output scales without adding editors.
4. iteration is cheap. feedback is simply a matter of prompting a few lines now, so we can test more versions and keep what lands.
agentic editing is the worst it'll ever be today, and it's already running rings around traditional editors/NLEs. content people who learn to direct it now will be miles ahead.
The hesitation around USD stablecoin rails is practical.
The on-ramp rate rarely matches the xe.com benchmark their balance sheets and SOPs run on.
Local stablecoins plus @RatioFX: a near-zero on-ramp and swaps at traditional FX rates, says founder & CEO @Iam_JohnCho.
all done in opus 5.5 and lemo opuscar btw, voiced with @ElevenLabs
stu (스튜) retweeted
Korea has some of the fastest infrastructure and most active crypto markets in the world, yet the won still can’t move onchain.
Watch the first ever episode of Won For All Series!
It breaks down why that matters and what it takes to fix it. 👇
stu (스튜) retweeted
저희 팀이 원화 스테이블코인을 다룬 4부작 영상 'Won For All'을 만들었습니다. 첫 편 '끊어진 레일'을 공유합니다.
한국은 디지털 자산에 가장 익숙한 나라 중 하나지만, 원화가 움직일 레일은 아직 없습니다. 저는 발행이 출발선이라고 생각합니다. 그다음 질문은 그 원화 스테이블코인이 실제로 쓰이느냐입니다.
앞으로 세 편에 걸쳐 그 레일을 어떻게 놓을 수 있는지 함께 보시면 좋겠습니다.
all eyes are on korea right now as stablecoin and digital asset legislation takes shape.
DABA가 다가옵니다. FSC가 이달 중 정부안을 공개할 예정입니다.
and then you have the issuers, with the spoils likely going to the banks rather than fintechs, although the holding ratios are still up for debate.
it's against this backdrop that we're releasing this four-part series titled 'Won For All'; a story of what korea needs to defend its currency sovereignty, and the cost of inaction.
원스코의 레일 이번에는 우리가 직접 깔아야 합니다 @KaiaChain.
the biggest problem facing south korea right now is not the KOSPI falling off a cliff
it is AI slop cafe music
6
14
763
korea's crypto tax just slipped (again) to 2029. some context for anyone outside korea:
the plan is 22% on yearly crypto profits above ₩2.5M (about $1,800). it was supposed to start in january 2027, and it has now been delayed four times.
two things point to 2029 as a political date:
korea has a general election in april 2028, and taxing millions of retail traders right before it would cost votes. korea also cancelled a planned tax on stock profits in late 2024 after investor pushback, so there's already a precedent for backing down.
the other reason is practical. CARF is a global agreement where countries share crypto exchange account data with each other's tax offices.
korea and the US start swapping that data in 2029. until then, korea can't easily see profits made on overseas exchanges, so taxing only local exchanges would push people offshore.
so 2029 gives the tax office full visibility, and it pushes the fight past the election.
whether it actually starts then is another question entirely.
the easiest way to go viral on x,
is to publish content about going viral on x.
the reason for this is simple: everyone on x wants to grow on x.
it's the one niche every single user is already in.
look at @ryangrowthh's article below, for example. his feed is the algorithm explained, the algorithm explained in depth, 30 creator tools, a free growth guide, and a podcast about x growth. value, delivered.
the algorithm breakdown did ~410k views and 8k bookmarks. posts on other topics do a fraction of that.
it's genuinely useful stuff. but it's also content about content, posted on the platform the content is about.
and it works. it'll probably keep working.
so if you want reach, talk about reach.
if you want to be known for something else, you'll have to accept a slower climb.
1
10
439
after all the comparison shots I’ve seen,
yes, it looks like they do bypass the iPhone imaging pipeline, but they actually don’t add that much to the processing.
the images are distinctly VSCO-esque; an unmistakable S-curve baked in, all hdr killed off, some split toning to give shadows and highlights a distinct tint.
if anything, I’d say the images that I’ve seen from this app are closer to those of an iPhone 7/8, which, if you are a bit of an anorak about iPhone imaging, you’ll know that’s exactly when iPhone cameras peaked.
would still rather shoot with my @FujifilmX_US than forking out $13 tho 🤷♂️
So I recently badgered my cinematographer @steveyedlin into making my ideal iPhone still camera app. It shoots RAW, bypasses the iPhone's look processing and instead uses the same image processing we use on our movies. apps.apple.com/app/apple-sto…
2
144
korea is in the middle of a metaverse renaissance rn
it all started with 산사, an app that emulates a temple, and has users meditating and finding peace amidst digital madness.
the dev counts a million users and increasing. and that's not the only metaverse that's gone viral.
within days there was:
- an online cathedral (inpeace)
- a sauna (사우나 고요)
- a 40-seat 독서실 where you just sit and study next to strangers
- a riverside 불멍 campfire
all open from a single link, most built on three.js, no login, no wallet, no chat. and some run purely on kakaopay tips.
in a country full of loners it makes sense. for people who grew up studying side by side in 독서실 and gaming in PC bangs, these apps give you company without asking anything of you.
the metaverse never died. it multiplied.
1
13
500
to expand on this:
there are two distinct ways of agentic editing emerging – headless (i.e. directly in the LLM) vs NLE (in an editor timeline)
professionals would be well-advised to get to grips with using the Davinci MCP to edit on their NLE, even if just to get rote tasks done, that's how much time it saves.
'normies' (for lack of a better term) who want to pump out videos at scale are better off sticking to headless editing as you can get 90% of the way there without worrying too much about production quality. save your looks/workflows as dedicated skills and you can pretty much one-shot an edit immediately.
what I have noticed though is that agentic NLE editing is not that intuitive yet e.g. claude insists on burning in subtitles as a .mov layer in davinci resolve rather than using the native text+ captions. it will also create custom assets for animations rather than use fusion which can take ages, especially on older machines.
still very bullish on agentic editing overall though, this is one of the most exciting developments in a while in the content space.
2
7
225
stu (스튜) retweeted
one week of agentic editing later
agentic editing is amazing, don’t get me wrong.
but after a few months it will go the way of vibe-coding — the supposed next goldmine of AI income will just be dominated by visual tells that scream slop.
right now the average person has no idea how to edit agentically, which is why you’re seeing creators shill their courses and cash in on this gold rush.
personally I’ve used claude to post daily news videos for the past week, and this is something that would have taken me eons longer in the past.
if you were creating content before agentic editing, congrats, your output just got a huge boost.
but if you think agentic editing is a panacea for growth while your core content processes are broken, you’ll be left shipping the same slop as everyone else around you.
p.s. there is an easy way to get past most of the hallmarks of claude-driven editing, which I'll be sharing very soon.