Jeff Yan 談 Hyperliquid：Primitive、AI代理人與監管新局

0:00回顧一年：Hyperliquid 的成長與突破

0:00大家好，我是 Hart Lambur。我是 Across 的創辦人，這是一個跨鏈橋接協議。我們和 Hyperliquid 合作，也是 Hyperliquid 生態系的一部分。很高興去年就和 Jeff 一起做過這場對談，幾乎就是一年前的今天。我想我上次表現還不錯，Jeff 又邀請我再來一次，希望今年能超越去年。Alright. So I'm Hart Lambur. I'm the founder of a cross-chain bridging protocol. We work with Hyperliquid and are part of the Hyperliquid ecosystem. I had the pleasure of doing this talk with Jeff about a year ago, almost exactly a year ago. And I guess I did well enough that Jeff invited me to do it again. Let's see if we can top last year.

0:33是啊，我們要努力讓今年比去年更精彩，這也是我們今天要聊的主題。其實我根本不用介紹 Jeff，大家都知道他是 Hyperliquid 的創辦人。Jeff，我覺得這一年真的很不平凡。我有一些數據可以分享，但在那之前，想問你回顧去年，有什麼讓你特別自豪的地方？你覺得哪些事情真的做對了？Yeah. Yeah, we have to try to top last year. And that's what we're going to talk about. To start, I don't really need to introduce Jeff. You all know, the founder of Hyperliquid. Jeff, I think it's been a pretty extraordinary year. I have some stats I can share, but before I get into that, when you look back at last year, what are you proud of? What do you think really worked?

1:02對，我們會再重溫很多重點，算是複習一下。但我覺得核心理念從去年到今年其實沒什麼改變。對我來說，最明顯的就是一切都變得更具體、更有感，大家的熱情也很明顯。用鏈上技術來提升現有金融系統的技術堆疊，這點很關鍵，而 Hyperliquid 算是站在這個中心。能夠以這種方式打造產品，讓更多開發者、機構和用戶都有機會參與，真的很令人興奮。Yeah, I think we'll go through a lot of the points again, kind of a repeat. But I think the core ideas haven't changed from last year to this year. For me, what stands out is how much more tangible everything feels. The excitement all around. Improving the tech stack of the existing financial system using on-chain technology, right? And Hyperliquid is kind of at the center of that. And it's been a pleasure to build in a way that opens up opportunities for a large group of developers, institutions, and users. That's really exciting.

1:44我記得去年我們談的是願景，但今年這個願景真的實現了，雖然很難相信。去年我們在這裡時，HIP 3 剛宣布，實際上是我們對談後一兩週才正式上線。那時 Hyperliquid 幾乎只被認為是做加密永續合約的平台，但現在情況完全不同了。從數據來看，月交易量跟去年差不多，但要知道，BTC 和 ETH 的價格其實跌了 30% 到 40%，所以我們實際表現其實更好。I think last year we talked about this as a vision, but I feel like over the course of this year, even though it's hard to believe, it really came to life. When we were here last year, HIP 3 was announced, but I think it only went live a week or two after our talk. Last year, when we spoke, Hyperliquid was almost only known for crypto perpetuals. To address that, that's just not true anymore. Statistically, the monthly trading volume I checked is about the same as last year. But to put that in perspective, Bitcoin and ETH prices are down 30 to 40 percent. So we've actually performed much better on an adjusted basis.

2:25但更重要的是，活動分布的品質完全不同了。我們現在有 HIP-3 市場、RWA（實體資產）、實體資產永續合約，這些大概占了總交易量的 30% 到 50%。我們現在還能產生 USDC 收益。提醒大家，去年我們對談時，所有存放在 Hyperliquid 的 USDC 都沒有為協議帶來收益，現在則有了。第一筆收益大約是上週進來的，金額是 1,500 萬美元，5 美元。如果年化，大概是 1.77 億美元，這對 Hyperliquid 是額外的收入。But the quality of the distribution of activity is fundamentally different. We now have HIP-3 markets, Real World Assets, Real World Perps, and we'll talk more about that, but they make up about 30 to 50 percent of the total volume. And we're now also generating USDC yield. For the audience, just to make sure everyone remembers: a year ago, when we spoke, all the USDC in Hyperliquid generated no yield for the protocol. Now it does. The first payment came, I think, last week: $15 million, $5. Annualized, that's $177 million, so additional revenue for Hyperliquid.

3:24回頭看協議收入的品質，你會預想到協議在這一年內能有這麼多元的成長嗎？Jeff 回答說，其實我們並不是特別追求收入，這只是為用戶創造價值的副產品。我們的指引原則，至少在 Labs 和整個生態系，就是要做出很酷的東西，不只是現有東西的小改版，而是要做出全新的東西、帶來新典範。像你剛提到 HIP 3，從想法到實現，這就是我們年底回顧時會覺得超酷的事。So when you look back and consider the quality of protocol revenue, could you have foreseen that the protocol would diversify so much within a year? Jeff: Yeah, well, I guess we're not really optimizing for revenue. That's more of a byproduct of providing value to users. But I think our guiding star, at least at Labs and in the ecosystem in general, is to build cool things. Not things that are just slight variations of what's already out there, but we prefer to do something for the first time and really introduce new paradigms. I think the first thing you mentioned with HIP 3, the journey from idea to reality, that's the kind of headline we look back on at the end of the year and say, wow, that's super cool.

4:13USDC 創新與生態系力量

4:13回到 USDC 這點，我認為 Coinbase 和 Circle 一起作為 CCPV2 的部署者，是很獨特的創新。真的要給這兩家公司掌聲，他們投入了很多努力才讓這一切成真，也展現出他們真的很重視 Hyperliquid 生態系的用戶。我們也非常尊敬他們，他們為了讓 USDC 合規、成為穩定幣，做了很多工作。我覺得這是一個完美的協同，這也是我們 Labs 團隊很自豪的亮點，就是打造能夠創造新可能性的技術。And to your point about USDC, I think Coinbase and Circle together taking a step forward and acting as deployers of CCPV2 is a unique innovation. Hats off to both companies, right? They've put in so much work to make this possible. And I think it shows that they really value Hyperliquid users in the ecosystem. We respect them a lot. They obviously do a lot of work to keep USDC compliant and make it a stablecoin. I think that's an absolutely perfect synergy. That's the kind of headline we're proud of at Labs—building technology that makes new things possible.

4:59整個金融體系，去年我們就討論過，Hyperliquid 團隊其實只有十一、十二個人，現在可能多一點點，但我們對成長很謹慎。至少在 Labs，我們寧願少招一點人，但每個人都能產生更大影響，這樣更有趣，也讓大家都很有使命感。你去年也強調過，其實不能只說我們只有這十二個人，因為現在更是如此。你怎麼看？The entire financial system—one of the points we discussed last year—is that it's still a team of eleven, maybe now twelve people. It's a little bigger now, yeah. We're very selective about growth. At Labs, at least, we prefer to hire fewer people who can move more. We think it's more fun that way, and everyone really feels committed to the vision. And a point you insisted on last year was: well, it's not fair to say we're just these twelve people. Because that's even truer today, I think. Explain that. What do you mean?

5:38回顧去年到今年，對我來說，最重要的改變是大家現在都認真看待這件事。這不只是個噱頭，而是真正能用的技術，是中立的，可以幫助企業實現目標、為用戶創造更多價值。其實，生態系就是所有在這個協議上建構的人們的總和。所以我認為，雖然去年這點已經很明顯，今年更是如此。Looking back from last year to this year, for me, the most important thing is that people now take this seriously. It's not just a gimmick. It's real technology you can use. It's neutral and helps you do things as a business to provide more value to your users. And I think, when you look at it that way, the ecosystem is just the sum of all the people building on the protocol. So I think, even though that was already true last year, it's even more true this year.

6:13Hyperliquid 對傳統金融的影響

6:13這個想法絕對不是噱頭。我想再談談永續合約產品。因為我們現在在舞台上，值得一提的是，Hyperliquid 所做的事情，已經明顯影響到傳統金融市場的運作。像 CME 開始往 24/7 市場發展，股票市場也逐漸提供 24/7 交易機會，傳統交易所也開始推出新的永續合約產品。我們等等還會再聊，但像 Robinhood、Kalshi 這些偏 Fintech 的公司也都在推永續合約。當這些大公司都受到你們帶動的創新影響，就很明顯這不是噱頭。So this idea is definitely not just a gimmick. I want to revisit the perp product. Since we're on stage, it's worth mentioning that what Hyperliquid has done has clearly influenced how traditional financial markets operate. The CME is moving toward 24/7 markets. Stock markets are evolving to offer 24/7 trading opportunities. We're seeing traditional exchanges launch new perp products. We'll come back to this, but traditional fintech-oriented companies like Robinhood and Kalshi are also launching perp products. It's obvious it's not just a gimmick when all these huge players are picking up on the innovations you all have driven.

7:26這比較像評論不是問題，但你對於自己有這樣的影響力有什麼感覺？Jeff 說，他不覺得大家是被 Hyperliquid 影響才這麼做。如果一個想法夠好，它會自己說話，首先會吸引用戶。用戶會說，這正是我想要的，讓我能更好地表達對價格的看法，不用再煩惱要選哪個合約、哪個履約價。現在有更純粹的方式來表達自己的觀點，很多用戶覺得這是從零到一的轉變，也真的採用了這種方式。That was more of a comment than a question, but how does it feel to have that kind of influence? Jeff: I mean, I don't think people are influenced by us. If ideas are good, they speak for themselves, and I think they speak to users first, right? Users say, this is exactly what I want to do. It lets me express my view on the price of something better, because I don't have to think about which contract to trade, which strike price, etc. Whatever you did before, now you have a much purer way to express your view. And I think a lot of users see that as a zero-to-one change and are making that shift.

7:58這個想法也吸引了你剛提到的其他優秀開發者。我很高興能成為其中一份子，但我不會說大家都在看 Hyperliquid 然後想著要抄襲我們的東西。更像是這些想法本身就很有潛力和價值。對，我們這幾年一直這麼說。那我們再重申一次，去年我們討論過、也很有共識的一點，就是永續合約產品的品質。This idea also speaks to all the other great developers you mentioned. I think it's cool to be part of that. But I wouldn't say people are looking at Hyperliquid and thinking, oh, we just have to copy whatever they're doing. It's more like these ideas themselves have a lot of potential and merit. Yeah, that's true. We've been saying that for years. Let's repeat that again. One of the things we talked about last year and really agreed on is the quality of the perps product.

8:34永續合約的優勢與市場演變

8:34為什麼永續合約這麼好用？也許對在場的大家來說很明顯，但其實永續合約是一個非常漂亮的產品。去年我們都說過，BitMEX 剛出現時就吸引了我們的注意，讓人很好奇它怎麼運作。Jeff，你為什麼覺得永續合約這麼優雅？為什麼現在這麼多傳統大公司都在模仿？Why do perps work so well? Maybe it's obvious to everyone here, but I don't think it's obvious that perps are such a beautiful product. Last year, we both talked about how BitMEX caught our attention when we first saw it. It was like, wow, this is different. You were curious how it worked. But to repeat for the audience, Jeff, why do you think perps are so elegant? Why do you think all these other, more traditional players are copying them now?

9:07這其實就是我剛才講的重點。基本上，永續合約讓每個資產都有一個標準的交易市場。我覺得這是一個很有力的想法，因為它沒有到期日，用戶也不用做太多決策。如果用戶需要做很多決策，做市商也要考慮各種可能，這會讓流動性分散在不同市場，雖然其實大家都是在交易同一個標的。當然，有些複雜的金融工具有其存在理由，但大部分情況下，大家只是想對某個數字表達看法，這時永續合約就很適合。Yeah, that's kind of the point I made earlier. Basically, with a perp, for every asset, there's a canonical market to trade. I think that's just a very powerful idea, that it doesn't expire and you don't have to make decisions. Because if users have to make decisions, market makers also have to consider all the different decisions users might make. That naturally fragments liquidity between different markets, which ultimately all express the same underlying asset. We can talk about it later, but there are real reasons for these kinds of convex instruments. But I think, for the most part, if you just want to express a view on a number, with a kind of linear payout—which I think is the majority of use cases crypto perps have found—I think that translates very well to traditional finance.

10:01我認為，價格發現的方式其實不會因為資產類型不同而改變。這就是一個好點子，所以你會得到一個更優秀的產品。去年你說過一句我很認同的話，就是永續合約這個產品，我們已經證明它可行，數學模型也成立。跟期權或期貨那種要一直換合約、換到期日的產品不同，永續合約把流動性集中在一個數字上。其實很多傳統金融的本質，就是預測一個數字，我覺得這真的很對。I think the way price discovery works doesn't really change depending on the asset. It's just a good idea, and I think in this case you just get a superior product. One thing you said last year that I found very precise is, okay, the perp product, we've proven it works. The mathematical formulation works. And unlike options or futures with rolling expiries and all that stuff, you have a number. You concentrate liquidity on a number, and you said a lot of traditional finance comes down to predicting a number. And I think that's really true.

10:42我們去年預測過，永續合約會延伸到實體資產或股票永續，這點也真的實現了。我想再回到 HIP-3，聊聊你覺得哪些地方做得好、你最自豪的是什麼。永續合約是一個很棒的產品，因為它就是讓大家預測一個數字，去年我們預測油價、Pre-IPO 永續合約，這些都很準確。Jeff，回到 HIP 3，你覺得還有哪些地方做得特別好？One of the predictions we made was that perps on real world assets or stock perps would be a big deal last year. And we actually nailed that pretty well. I want to come back to HIP-3 and ask what you think worked well and what you're proud of. The observation that perps are a great product for predicting a number seems to have been very true last year when we had to predict oil prices or pre-IPO perps. So, yeah Jeff, back to HIP 3. What else can you say about why you think it worked so well?

11:29HIP-3 創新與用戶賦能

11:29我覺得這像是很多因素同時到位。對我來說，最令人興奮的是過去一年 HIP-3 市場的創新爆發，這些市場的共同點就是讓用戶能做以前做不到的事。這點讓我很有感，因為金融常常被認為很複雜，只有有優勢的人才能賺更多，但 HIP-3 讓更多人有機會參與。Yeah, I think it's like the stars aligned on a lot of different factors. For me, one of the most exciting things about the explosion of innovative HIP-3 markets introduced in the past year or so is that the common theme of what works well is that it enables users to do something they couldn't do before. And that really speaks to me, because I think finance is often seen as something that's supposed to be very complicated and lets people with some advantage over others profit more.

12:11金融系統的願景與公平性

12:11感覺世界上很多人，甚至可能是大多數人，都對金融系統有一樣的看法。我認為這其實跟金融的本質是相反的，對吧？重點應該是金融系統要對每個人都開放，應該是一個讓大家都能參與資產累積的機會，最好是讓地球上每個人都能參與。It feels like a lot of people in the world, maybe even the majority, see the financial system the same way. And I think that's actually the opposite of what it's really about, right? The point is that the financial system should be accessible to everyone. It should be a means of wealth creation that's as broadly accessible as possible. If possible, for everyone on the planet.

12:32我覺得如果你去問經濟學家或這個產業裡的人，問他們為什麼做這份工作、為什麼熱愛這份工作，最終大概也是因為這個原因。他們覺得自己在為這個大願景盡一份力，讓人類社會隨著時間進步。I think if you talk to economists or anyone in the industry and really ask them why they do what they do, why they love their work, it's probably ultimately because of this. It's like they're contributing a small part to this big vision, that humanity improves over time.

12:51回到 HIP 3 這個主題。不管你剛剛提到的是什麼資產，如果只有少數人在大部分成長期間能交易，等到成長紅利都被少數人拿走後，才開放給大眾，這種模式根本不可持續。我不覺得這是有意為之，但這是經濟發展過程中的副作用。And when I come back to HIP 3. Whether it's, as you mentioned, an asset that only a few people can trade during most of its growth and only becomes accessible to the public after the growth has already been realized by a few privileged people. I think that model is just not sustainable. I don't think it's intentional, but it's a byproduct of how things change in the economy.

13:19我認為，作為一個協議，應該成為一股平衡的力量，告訴大家：只要你有創新想法，想讓更多人能接觸金融系統，你就可以用這個協議，因為它的精神就是這樣。你可以在這個中立的協議上打造你的應用，它不會擋你的路，反而會支持你，讓你能接觸到原本沒機會的用戶。這個故事對我來說很有意義，也要大大稱讚那些真正努力實現這一切的建設者。And I think, as a protocol, to be a balancing force and say, hey, anyone who has an innovative idea for expanding access to the financial system for others, you also have this protocol that's aligned with that ethos. You can come and build on this neutral protocol and it won't get in your way. It will support you, and you can then reach users who now have opportunities they didn't have before. That story is really meaningful to me, and of course, big kudos to the builders themselves who do the hard work to make it happen.

13:59把想法落實才是最難的部分，這對我來說是最重要的因素。當然還有你剛剛提到的其他事，比如 24/7 可以交易現有資產，像永續合約那樣，流動性集中、價格發現更容易。我覺得這些都是在開放金融市場的核心價值上，進一步放大的功能。Connecting the idea with execution, I think that's the hard part. For me, that's the most important factor, but then there are the other things you mentioned. Like being able to trade an asset 24/7 in a kind of perpetual format, where liquidity is pooled and price discovery is easier. I think these are all multiplicative functions that build on the core utility of open financial markets.

14:24DeFi 的全球普及與創新

14:24你剛剛講的時候，我突然想到一個月前我還翻出 2018 年的舊簡報給團隊看，那時還沒發生新冠疫情。當時我對 DeFi 最興奮的兩件事，就是全球市場的普及性和降低金融創新的門檻。現在這些真的都在發生，真的很酷。While you were speaking, it occurred to me that about a month ago I dug up an old slide for my team from 2018, so before Covid. The two things that excited me about DeFi back then were universal market access and lower barriers to financial innovation. And I think it's damn cool that that's literally happening now.

14:51回到主題，Hyperliquid 是全球都能用的。我們等下會再聊 HIP 3，但 Hyperliquid 的目標就是全球普及、整合所有金融、降低金融創新的門檻。這正是 HIP 3 做到的，現在我們 24 小時有股票市場、IPO 永續合約、週末還能交易石油，這只是開始。And back to the topic, okay, Hyperliquid is globally accessible. We'll talk more about HIP 3 later, but Hyperliquid is meant to be globally accessible, unite all finance, and lower the barriers to financial innovation. That's exactly what HIP 3 has achieved, where we now have 24/7 stock markets, IPO perpetuals, oil trading on weekends. And that's just the beginning.

15:18能夠提供這種高品質的金融創新，不是玩票性質，而是真正大規模運作的東西。我現在有點激動，因為這正是十年前很多人投入 DeFi 的願景，現在真的大規模實現了，真的很酷。And being able to offer this level of financial innovation at the highest quality, where it's not gimmicks but things that work at scale. I'm kind of gushing right now because I think this was the vision of many people who came to DeFi a decade ago, and it's being realized at scale right now. So that's really cool.

15:40跟大家分享一個數據：今年七月，Hyperliquid 有 51% 的交易量來自 HIP-3 市場。這麼短時間內，沒人能預料到吧？And to share this with the audience: in July, a full 51 percent of Hyperliquid's volume came from HIP-3 markets. You couldn't have predicted that in such a short time, right?

15:54產品市場契合與信任建立

15:54對啊，我們其實不太做預測，因為很多事都超出個人、團隊甚至整個產業的掌控。我不認為這是必然的結果。當然，51% 這個數字很大一部分來自一個很厲害的團隊 Trade XYZ，他們真的把我們剛剛討論的那些創新落實了。Yeah, we try not to make predictions, just because so much is outside the control of a person, a team, or even the industry. I don't think this was a guaranteed outcome. Of course, if you look at the 51 percent, a large part of that came from a very outstanding team, namely Trade XYZ, which really implemented a lot of the innovations we talked about earlier.

16:29但我一直相信，光有想法其實沒什麼價值，真正困難的是執行。像你剛剛說的，七月時超過一半的交易量都來自這裡，這證明產品市場契合度有多高，也顯示原本的金融產品和新產品之間其實有很大落差，大家願意轉換。But I firmly believe that ideas themselves aren't worth much. It's the execution that's the hard part. So, as you said, in July, more than the majority of the volume came from this, and to me, that speaks to how big the product-market fit was and how big the gap actually was, that people were willing to switch from their existing financial products to something new.

16:54我覺得，為了讓用戶願意一次又一次地給新產品機會，必須建立很高的信任，這也說明了開發者在執行上的品質。I think the level of trust that needs to be built for users to take that leap of faith again and again speaks to the quality of execution by the developers in general.

17:11Builder Codes 與金融基礎建設

17:11沒錯。接下來聊聊 Builder Codes，我們之前有提過，但我現在對背後的願景更清楚了。在我問你細節前，你要不要先簡單跟大家介紹一下 Builder Codes 的概念和運作方式？Absolutely. Let's talk about Builder Codes. We touched on it briefly before, but I think the vision behind it is clearer to me now. Before I bombard you with questions, do you want to briefly explain the concept of Builder Codes to the audience and how it works?

17:30這個想法其實很簡單。基本上，全球金融體系有兩種可能的願景。一種是公司把所有流程垂直整合，從清算、用戶進場、定義市場、到結算全部包辦，打造一個很棒的產品。這是過去 FinTech 或 Web2 很擅長的模式。Yeah, the idea is pretty simple. There are basically two possible visions for how global finance should be structured. One is, a company just delivers good execution and verticalizes everything from clearing, onboarding users, market definitions, to settlements if needed. So just verticalize the whole stack and build a really good product. That's what FinTech or Web2 have done very well so far.

18:07但我認為 Builder Codes 屬於另一種更激進的願景，比較像 DeFi，強調組合性。這個想法其實來自 DeFi，就像 Linux 或 Unix 的作業系統哲學，每個人都能把自己的部分做到最好。如果協議或架構允許不同的基礎元件無縫組合，最後就能形成一個不是由單一中心化實體擁有和營運的系統，而是由很多小但優秀的組件組合而成，就像樂高積木一樣。But I think Builder Codes are part of the other vision, which is maybe a bit more radical. It's more DeFi, a kind of composability. The idea actually comes from DeFi, as I said earlier. But it's similar to the Linux or Unix philosophy for operating systems. It's that different people can build their own parts really well, and if the protocol or architecture allows seamless composition of different primitives, you end up with a system that's not vertically owned and operated by a central entity, but rather an emergent property from a set of smaller but well-built parts, coming together like Lego blocks to form a structure.

18:55這就是我們的願景，Hyperliquid 非常相信這種模式。我們認為這樣的金融系統更強韌、更有彈性，最終也更公平、對用戶更好。如果是建立在公平、開放的協議上，大家都能在上面創新，這才是最棒的。I think that's the vision, and at Hyperliquid we believe very strongly in this latter vision. We think the financial system is stronger, more resilient, and ultimately probably fairer and better for users if it's built with the vision of a fair, open protocol that people can build on.

19:21Builder Codes 只是這種系統的一個面向，讓不同開發者可以打造自己的創新介面，無論是手機 App、專業機構級終端機，或是介於兩者之間的東西。你專注於你獨特的推廣方式，Builder Codes 讓你能在協議的交易層收取經過你整合的流量費用，完全自動化，無需談判。很多開發者覺得這種方式更自由，因為你可以直接接上協議，省下很多重複造輪子的工作，像是不用再寫一次訂單簿或清算邏輯，這些都不是創新的地方，何不直接用現成的基礎建設，專注在你獨特的創新？這就是我給開發者的建議。Builder Codes are just one aspect of such a system, where different developers can build their own innovative interfaces, whether it's a mobile app, a professional institutional terminal, or something in between. Whatever your unique distribution mechanism is, you focus on that as a developer, and Builder Codes let you collect fees at the transaction layer of the protocol for the flow that goes through your integration. It's completely straightforward, no negotiations. Many developers find this a much more liberating way to build, especially since you benefit from plugging into something that autonomously handles all the boring work, right? Nobody wants to write an order book again. That's pretty hard. It's hard to write clearing logic. There's a lot in there, and that's not where innovation happens. So why not use the proven infrastructure and focus on what's unique to you? That would be my pitch to developers.

20:33金融基礎建設的兩種策略

20:33這幾年我對這個願景又更有感，因為我們看到很多偏 FinTech 的公司出現。像 Robinhood 或 Calci，他們也有做永續合約，但是很垂直整合的方式，從上到下都掌控用戶體驗。Yeah, and I think this vision has become much clearer to me again in recent years, as we've seen these other more Fintech-oriented companies. Take Robinhood or Calci, for example. They offer perps, but in a very vertically integrated way. They own the user across the whole stack.

20:53如果要具體想像，Hyperliquid 用 Builder Codes 打造出一層基礎建設，現在有像 Velo 這樣的前端，還有 Valor、MetaMask、Phantom、FOMO 等等，這些公司都不用自己重做底層基礎建設，只要接上這個頂級的 Unix 式元件就好。我覺得這兩種策略的發展很有趣，其實某種程度上是互補的。To really visualize it: Hyperliquid has this infrastructure layer with Builder Codes. We now have frontends like Velo, which we should actually talk about, Valor, MetaMask, Phantom, FOMO, all these companies don't have to build the underlying infrastructure themselves. They just plug into this top-notch Unix primitive that does this part really well. I find it really interesting to see how these two strategies develop, you know? And they're kind of complementary in a way.

21:29Velo 與 Builder Code 策略的實例

21:29Jeff，我們來聊聊 Velo，這是一個很棒也很有趣的例子。Velo 原本可能被視為 Hyperliquid 的競爭對手，但他們其實採用了 Builder Code 的理念。這是第一個這樣的整合案例，要給 Velo 團隊大大的肯定。我希望這只是未來很多案例的開始。你可以把它當成競爭，但我覺得這不是正確的看法。But yeah, Jeff, let's talk about Velo as a great and pretty fascinating example of someone who could be considered a competitor to Hyperliquid, but actually adopts this Builder Code philosophy. Yeah, no, that's a great example. It was a first-of-its-kind integration, so big kudos to the Velo team for pulling that off. I think, hopefully, it's the first of many such examples. You can look at it as a kind of competition, but I don't think that's the right way to see it.

22:11如果你把 Hyperliquid 當作一個協議，它其實不是要跟你剛剛提到的那些公司競爭，也不是要跟任何公司競爭。它只是個工具、一層基礎建設，讓公司可以直接用。沒有人會跟網路競爭吧？大家就是用網路而已。希望未來會有更多這樣的例子，大家發現可以直接接上最有流動性的資產市場，像 Bitcoin，完全不影響原本的用戶流程，還能維持對用戶的掌控。當大家意識到這點，其實就很自然會採用，但這還需要建立信任，協議在這方面已經有進展，但還有很長的路要走。If you look at Hyperliquid as what the protocol really is, it's not trying to compete with anyone you mentioned, or any company at all. It's really just meant as a tool, like a layer that companies can take. Nobody competes with the internet, right? You just use the internet. So I think, hopefully, there will be more examples in the future that recognize this and get familiar with the technology, and realize, oh wow, you can instantly access the most liquid markets for assets like Bitcoin, things users obviously want to trade, in a way that doesn't disrupt the existing user flow. You can still maintain control over them. I think when that realization hits, it's kind of a no-brainer, but it also requires a kind of trust component, and the protocol is making progress there, but of course, there's still a long way to go.

23:10全球金融信任與在地品牌

23:10沒有人會跟網路競爭，這句話真的很棒，希望大家也能在 Twitter 上分享。另一個我這一兩年常跟你聊的想法是，金融領域的信任到底怎麼建立？其實從來沒有一個真正的全球金融服務品牌，金融服務一直都是在地或區域性的。雖然加密產業有試圖打造全球品牌，像 Binance、Coinbase，現在還有 Robinhood，但如果把 Hyperliquid 當作基礎建設，反而能讓像南非的 Bala 這種在地品牌建立與當地用戶的信任關係，這對金融運作來說很重要。I think nobody competes with the internet is a great quote. I hope people post that on Twitter too. Yeah, man. I find it extremely insightful. The other thought is something I've talked with you about a bit over the last year or two. It's interesting to think about how trust is built in finance. And I want to say, one observation is, there's actually never been a global financial services brand. Financial services have always been local or regional. We've had attempts from crypto to build global financial brands, like Binance or Coinbase, and now we have things like Robinhood. But what I find really interesting is, if you look at Hyperliquid as this infrastructure layer, it creates the possibility for local brands like Bala in South Africa to build the trust relationship with users in that region. And that seems important for how finance works. Yeah.

24:25未來一年與AI代理人展望

24:25如果我們展望明年，有沒有什麼你希望在 Hyperliquid 生態系推動或看到的？還是你其實只是來觀察？我覺得這個想法很簡單，沒什麼好多說的，重點就是執行。我認為應該要釋放更多潛力，就像你剛剛說的那樣。And I guess, if we look to next year, is there something you want to push forward or see in the Hyperliquid ecosystem? Or is it more about what you're looking for here? I think the idea is so simple. I don't think there's much more to say about it. I think it's just about execution. I think it should unlock more, like you said.

24:52我非常相信，連結金融體系最創新的方法還沒被發明出來。我希望明年能看到這樣的突破。看到大家的創意總是很酷。你有什麼預測或想法嗎？Yeah, I firmly believe that the most innovative ways to connect to the financial system still need to be invented. And I hope that happens next year. And it's always super cool to see what people come up with. Do you have any predictions or thoughts on that?

25:11我可以預測，這會是我們目前還無法預見的東西。雖然這預測有點無聊，但我預測我們會很難預測到底會是什麼。你覺得會有 AI 代理人或 AI 應用的成分嗎？Well, I can predict that it will be something we don't currently see as obvious. That's a pretty lame prediction, but I predict we'll have trouble predicting exactly what it is. Do you think there will be a component for AI agents or AI usage?

25:34我覺得這個方向超有趣。我的觀點是，大家可能對 AI 這個主題想太多了。很明顯，代理人在很多能力上已經超越人類，這個差距只會越來越大。我不認為代理人在技術上透過 API 整合進系統會有困難。真正的關鍵因素是，阻礙代理人全面普及的，可能只是因為它還不是一個 On-Chain 系統。Yeah, I think that direction is super interesting. My perspective is that people maybe overthink the AI topic. It's obvious that agents are already superhuman in many of their abilities, and that gap will keep growing. I don't think agents will have trouble integrating into systems via API from a technical perspective. The core factor that could block full adoption of agents is simply that it's not an on-chain system yet.

26:20對我來說，On-Chain 系統最重要的特性，就是資料和價值被放在同一個抽象層級。這不像舊的網路時代，只能傳遞位元組。現在你可以傳送請求和回應，但只要牽涉到金錢，就還是得找特定機構來決定你能不能做這件事。我覺得這種情況還是會發生，但會在不同層級，不是網路或協定層。For me, the main property of an on-chain system is that data and value are kept at the same level of abstraction. It's not like the old internet, where you could only push bytes back and forth. You can send requests and responses, but when it comes to money, you still have to go to these specific institutions that give you the okay or not. I think that will still happen, but on a different level, not at the internet or protocol level.

26:56一旦這個條件成立，我認為代理人就能完全接手，發揮他們的魔力。直覺上我很同意你說的。代理人和 On-Chain 之間有種共生感，缺乏這個會造成摩擦。不過我們還沒辦法預測這個新入口或新介面會長什麼樣子。And once that's the case, I think agents can just take the wheel and work their magic. Yeah, intuitively, I'm totally with you. There's something symbiotic between agents and on-chain, and the lack of it feels like a friction point. But as I said, I don't know if we have a prediction yet for what this new access point or interface will look like.

27:23我覺得這是很多公司都在努力的共同預測，就是你只要設定幾個高層次目標。我們剛剛聊到金融系統的目的是幫助人們累積財富、過更好的生活。我認為代理人會在這裡扮演輔助角色。你需要一個可接近的系統，也許有個超智慧體能幫你把目標轉化成具體操作，這些操作只有在這種系統裡才有意義，才能發揮全部潛力。也許明年我們可以回來檢討這些想法。I think it's a pretty common prediction that a lot of companies are working toward, that you just set some high-level goals. We talked about how the purpose of a financial system is to help people build wealth and live a better life. I think agents will probably play a complementary role in that. You need an accessible system, but maybe there's a task for a superintelligent being to help you turn your goals into concrete operations, which can only work in such a system. That's where the full potential comes in. Maybe we'll revisit this next year.

28:06HIP-4 與 Outcome Primitive 的願景

28:06我們換個話題來聊 HIP-4。我想確保有時間討論這個。大家如果有在關注 HIP-4，那是另一個 Hyperliquid Improvement Proposal，重點是讓結果型市場（Outcome Markets）可以在 Hyperliquid 的訂單簿撮合引擎裡運作。這功能五月時其實就很低調地上線了。不過很多人以為 HIP-4 只是 Hyperliquid 上的預測市場，但我認為這個 Primitive 遠遠不只如此。你怎麼看？Let's switch topics to HIP-4. One thing I want to make sure we have time for is, if people are looking at HIP-4, that's another Hyperliquid Improvement Proposal, and it's about enabling outcome markets—so binary and scalar markets—within the settlement infrastructure of the Hyperliquid orderbook matching engine. I think it was quietly introduced in May. But I think a lot of people just see HIP-4 as prediction markets on Hyperliquid. In my view, this primitive is much broader than that. What do you think?

28:49這也呼應我們之前討論過的主題，為什麼要有 Perps。我不認為答案是 Perps 就是萬能，也不是說 Perps 比其他東西都好。我認為金融體系真的需要凸性（Convexity）。凸性基本上就是說，當你想做預測或持倉表態時，你不希望獲利只是價格的線性函數，而是希望有下檔保護或更高的上檔機會。這就是凸性 Perps 的概念，而 Outcome Primitive 就是要補這個缺口。Yeah, that also ties back to our earlier point about why perps, and I don't think the answer was ever that perps are all you need or that perps are better than everything else. I think there's a real need for convexity in the financial system. Convexity basically means that when you want to make a prediction or take a position, you don't want the payout to be a linear function of price relative to the current price. You want some kind of downside limit or a higher upside probability. That's what convex payout perps are, and that's exactly the gap the outcome, HIP-4 primitive, fills.

29:33我認為 Hypercore 的願景是 Primitive 不要太多，但每一個 Primitive 都非常有用，能以多種方式發揮作用，並和現有 Primitive 產生很多協同效應。Outcome 是建立在現貨交易之上的，這很有趣，因為它讓很多功能可以共用。你剛剛說的沒錯，它可以是二元、連續或介於兩者之間。重點是它有凸性，還有一種無清算（No-Liquidation）的設計，這對某些市場來說很有吸引力。I think the vision of Hypercore in general is that there aren't too many primitives, but each one is very useful in many different ways and unlocks a lot of synergies with existing primitives. Outcomes are built on top of spot trading, which is pretty interesting because it enables a lot of shared functionalities. Like you said, they can be binary, scalar, or anything in between. The main idea is that they're convex, and they also have a kind of no-liquidation form factor, which can be very interesting for some of the markets we discussed earlier.

30:17我會把 Outcome Primitive 當成一個數學工具，讓大家可以創造新東西。願景不是說這個 Primitive 就是為了現有世界的某個用途，而是希望有人能用它做出我們想像不到的應用。Outcome SIP 4 就提供了這種機會，不管是金融、二元還是其他各種可能。I would really just see it as a tool, a mathematical tool, that lets people create new things. The vision isn't to look at something that already exists in the world and say, oh, that's exactly what this primitive is for. I think it's even cooler if someone can use a primitive in a way nobody could have imagined. Outcome SIP 4 specifically offers that great opportunity, whether it's financial, binary, or all these different possibilities.

30:51沒錯，這是一個 Primitive，而且是可替代的。你可以交易、轉移這些 Token，也都已經 Token 化，這很酷。我一直在想，如果把這和 Bildercode 結合，現在有人想建立自己的預測市場交易所，為什麼還要自己從頭做？你其實可以像租用 Hyperliquid 生態系一樣，像選擇權一樣。很多人問，為什麼 Hypercore 或 Hyperliquid Protocol 沒有選擇權？我認為這就是發展選擇權的基礎元素，重點是怎麼用這個設計做出直覺的選擇權體驗。Absolutely. It's a primitive, and they're also fungible. You can trade the tokens, send them, and they're also tokenized, which is pretty cool. I've always thought, if you combine this with Bildercode, and now someone wants to build their own prediction market exchange, why build a prediction market from scratch? You can basically rent the Hyperliquid ecosystem, like with options. A lot of people ask, hey, why aren't there options on Hypercore or the Hyperliquid protocol? I think this is the fundamental element to develop options. It's just a question of how you take this form factor and create an intuitive options experience for users.

31:37我覺得選擇權很難有很多原因，但我們在 Labs 做 HIP-4 這個基礎元素時，認為這是一個很好的折衷和平衡，能創造有流動性的選擇權市場。我可以多聊一點，但不確定大家有沒有興趣。I think options are difficult for many reasons, but from our perspective at Labs, working on HIP-4 as this fundamental element is a pretty good compromise, a balance of all aspects to create liquid options markets. I can talk more about it, but I don't know if that's interesting.

32:01我自己也是這方面的宅宅，但我完全同意，這就是建立選擇權市場的方式。你要在很棒的 One Delta 產品上建立深度，避險也在同一個協定上完成。這一切結合起來，讓選擇權和永續合約之間可以做 Portfolio Margin。我覺得這些都等著生態系團隊來實現。I mean, I'm a nerd about these things too, but I totally agree that's how you build options markets. You build this kind of depth on really great one-delta products. The hedging happens on the same protocol. All of this combined, portfolio margin between options and perpetuals. I think all of that is just waiting for teams in the ecosystem to build.

32:27監管動態與許可型市場

32:27我們時間不多了，我還想談一個重點。我們這裡不討論監管策略，這不是適合的場合。但我想評論一下現在監管進展的速度。我查了一下，去年 SEC 和 CFTC 發了 42 份直接針對 DeFi 和永續合約的信件或通知。大家想想，光是比特幣 ETF 就花了十年才核准。現在監管機關的速度真的很驚人。But we have little time left, and I want to make sure we hit another point. We can't talk about regulatory strategies here—this isn't the right place for that. But I want to comment on the speed things are moving on the regulatory front. I did my own research and found that in the last year, there were 42 letters or notices from the SEC or CFTC directly relevant to DeFi and perpetuals. To put that in perspective, it took a decade for the Bitcoin ETF to get approved. The speed regulators are moving at now is incredible.

33:22你們最近在測試網上推出了 HIP-3 Star Market，這是 HIP-3 的許可型版本。雖然我們不談監管策略，但我想知道你怎麼看許可型市場在 Hyperliquid 以及你們想打造完整金融體系的願景中扮演什麼角色？You recently launched the HIP-3 Star Market on testnet. For the audience, this is a permissioned version of HIP-3. Again, without going into regulatory strategy, I want to know how you see permissioned markets fitting into Hyperliquid and your vision of mapping all of finance here.

33:44這其實呼應我們前面談 Primitive 的討論。Primitive 應該是普遍有用的，但不一定每個人都要用。你不需要用到 N 個 Primitive 才能成功。這裡的想法是，開發者真的有需求，希望在現有 HIP-3 Primitive 上多一些控制權。這些額外控制權是給提供者用的，而且是選配，提供者可以決定是否適合自己的應用場景。Yeah, that ties back to our earlier discussion about primitives in general, right? These are things that should be generally useful, but they don't have to be useful for everyone. There's no requirement to use at least N primitives to build successfully. The idea here is simply that there's real demand from developers to have these additional controls over the existing HIP-3 primitive. So there are just some extra controls for providers, and it's optional—the provider decides if it makes sense for their use case.

34:17我認為許可型市場會開啟另一個需求族群。許可型不代表就沒有 On-Chain 金融的其他優點，也不是和協定互動的唯一方式。像我們剛剛提到的 Valor，我相信國際市場和區域性許可型市場的流動性可以共存。最終目標就是打造對最多開發者和提供者有用的技術。But I think permissioned markets ultimately unlock a different demand segment. And I believe the fact that markets are permissioned doesn't mean they don't also have the other great properties of on-chain finance, and it's not the only way to interact with the protocol. Like we mentioned Valor earlier, right? I believe liquidity between international markets and more regional permissioned markets can coexist. Ultimately, the goal is to build technology that's useful for as many developers and providers as possible.

34:54我很喜歡這個策略。就像你說的，這是 Unix 式 Primitive，可以加上額外控制，對受監管市場很有用。我們快沒時間了，但我想預測一下：去年我們預測 Equity Perps 會發展得不錯，我現在預測受監管的 Perps 會變得很大，不只是因為可能進入美國市場，而是機構能夠全面採用這種結構。Jeff，你有什麼最後的預測嗎？Yeah, and I like the strategy. Like you said, it's this Unix-like primitive that implements extra controls, which can be useful in regulated markets. We're almost out of time, but I want to make my prediction: last year we predicted equity perps would do well, and now my prediction is that regulated perps will get really big—not just because they might enter the US market, but because institutions will be able to fully adopt the perp structure. Jeff, any last words or predictions for next year?

35:38On-Chain 金融的未來與總結

35:38我的預測是，會有很強的動能。大家會意識到 On-Chain 技術的優點，重點不是要放棄現有企業，而是升級他們的技術，利用 Hyperliquid 和全球流動性網路。我預期我們會看到更多這樣的案例，回頭看會覺得不可思議這些事竟然一年內發生。這就是我的感覺，也是我模糊的預測。那我們就以 Hyperliquid 作結，謝謝大家來參加和收聽。Yeah, predictions. I guess my prediction would be a broader form of what you said—that there's going to be really strong momentum. I think people are recognizing the advantages of on-chain technology. It's not about abandoning existing businesses, but about upgrading their technology, building on Hyperliquid, and leveraging global liquidity networks. So my prediction is we'll see a lot more of that, and we'll look back and say, wow, crazy that all this happened in just a year. That's the vibe I have, and that's my vague prediction. So let's end it with Hyperliquid. Thanks everyone for coming and listening.