After Three Years at Multicoin, My Thoughts on Kyle the 'Shill King' Announcing His Exit

marsbitОпубликовано 2026-02-06Обновлено 2026-02-06

Введение

Kyle Samani, co-founder of Multicoin Capital, is stepping back from the firm, a move that reflects his strategic focus on pursuing asymmetric, non-consensual bets rather than managing a conventional fund. According to the author, a former Multicoin partner, Samani excelled at identifying and investing in high-conviction, contrarian theses—such as Solana’s potential as a "Nasdaq on-chain"—long before they gained mainstream acceptance. While some have questioned why Multicoin missed certain trends in the current cycle, like leading perpetual DEXs, the author explains that these areas had already become consensus plays, which did not align with Samani’s investment philosophy. His strength lay in early, non-consensus conviction. The author expresses confidence in Multicoin’s remaining team, noting their expertise in on-chain trading, AI, and regulatory strategy. Meanwhile, Samani is expected to explore new ventures, possibly in data-centric businesses. The piece concludes with a personal reflection: everyone has a unique “game” they are meant to play. For Samani, it was asymmetric investing; for the author, it’s building value-transfer protocols. The key is to focus on what one does best and play to win.

Author: mable.sol(Former Multicoin Capital Partner Mable Jiang)

Compiled by: Ken, Chaincatcher

Everyone can only excel at one type of game. You should spend half your time figuring out what game you're good at, and the other half fully focused on playing it well.

Kyle Samani knew exactly what his was.

His strategy was to first develop a theory, then find the right projects—and persistently wait for the theory to be validated (while occasionally assessing if the theory still held). In building a "Nasdaq on-chain" with Solana, he saw this thesis from the very beginning (even though Solana is still some distance from this vision today, that's not the point), and executed this strategy by leading multiple rounds of financing—even participating in discounted OTC rounds long after SOL's mainnet beta launch. He did everything he could to facilitate Solana's development.

His core gamble was in finding asymmetric opportunities, i.e., non-consensus bets. Haseeb was right: probably no one is better at this than Kyle.

People question why Multicoin didn't capture any of the leading perpetual DEXs or related infrastructure this cycle. No one knows better than I do that Multicoin was a firm believer in perpetual DEXs long before anything existed this cycle. Even before many in the industry understood how central limit order books worked—you'd be surprised, but it's true, after all many of us were originally cypherpunks who later adapted to become trading experts (laughs). Multicoin studied basically every perpetual DEX and analyzed the trade-offs. There were many blog posts about the design space and why certain designs were better—you can find them on multicoin.capital.

So why did they "miss" it? Because perpetual DEXs and on-chain liquidity—these are no longer non-consensus bets. Of course, they still have huge growth potential. But consensus bets hold no appeal for Kyle. That was never his game.

And for a fund, the goal is to make money for LPs. So, managing a fund by the book is no longer the game Kyle wants to focus on playing. Running a data company might be something he's interested in—this has never been successfully done before (if you're interested, you can probably listen to the podcast I did with him on FWRD). There are still many asymmetric opportunities in the world waiting to be discovered. Kyle knows his game, so he decided to focus on playing the game he's good at and interested in.

No one needs to worry about who is moving on to other fields. When you know what your game is, you should focus on playing it well and aim to win. Winning isn't the entire purpose, but devising a winning strategy is part of the game experience.

Many worry about what Multicoin will be like without Kyle. But I'm not worried, because I know the opportunities ahead for all of us align with the strengths of Tushar, Brian, Spencer, Shayon, Vishal, Matt, Greg, and the rest of the team: on-chain trading, AI, and compliance. As long as the team focuses on playing this game correctly, it will remain top-tier in the industry.

For me, I've been a builder for almost four years now. Looking back, my time at Multicoin was unique—intense debates with Kyle in investment committee meetings, strategic exchanges with Tushar, and exploring the design space of many things with every IC member. I know I can't be as good as any of them, but I've also found my passion and the field of play that belongs to me.

For Kyle, the Web3 dream has turned into nothing but cold ledgers. For me, in this highly commoditized world, these ledgers condense real value transfer. When all other filtering mechanisms completely fail, money becomes the world's only relatively scarce resource to express opinions—and in the near future,恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕连金钱本身也算不上了恐怕极有可能连金钱本身也算不上了。This is why I am building an asset generation protocol (@trendsdotfun).

He and I never needed to agree on what game to play. But we sincerely wish each other well and will cheer for each other from the sidelines.

You only live once, focus on the game you are best at.

I'm just playing games, but I play to win.

Take it easy, but don't slack off.

Связанные с этим вопросы

QWhat was Kyle Samani's core investment strategy at Multicoin Capital according to the author?

AKyle Samani's core strategy was to first develop a theory, then find projects that fit that theory, while persistently waiting for the theory to be validated. He focused on finding asymmetric opportunities and non-consensus bets, such as his early and continued investment in Solana to build a 'Nasdaq on-chain'.

QWhy did Multicoin Capital 'miss' investing in leading perpetual decentralized exchanges (DEXs) this cycle?

AMulticoin missed investing in leading perpetual DEXs because these opportunities were no longer non-consensus bets. While they had significant growth potential, Kyle was not attracted to consensus bets, as that was never his game. The fund had extensively researched perpetual DEXs but chose to focus on asymmetric, non-consensus opportunities instead.

QWhat are the new focus areas for Multicoin Capital after Kyle's departure, as mentioned by the author?

AThe new focus areas for Multicoin Capital are on-chain trading, artificial intelligence, and compliance. The author expresses confidence in the remaining team members, including Tushar, Brian, Spencer, Shayon, Vishal, Matt, and Greg, to excel in these areas.

QWhat personal project is the author, Mable Jiang, building and why?

AMable Jiang is building an asset generation protocol called @trendsdotfun. She believes that in a highly commoditized world, ledgers凝结了真实的价值传输 (condense real value transfer), and money is becoming the only relatively scarce resource to express opinions. Her project aims to address this.

QWhat is the overarching life and career advice the author gives based on her experience with Kyle?

AThe overarching advice is that everyone should spend half their time figuring out what game they are good at and the other half fully focused on playing that game well. The goal is not just to win, but formulating a winning strategy is part of the experience. One should focus on the game they are best at and play to win.

Похожее

Agents Take Over Traffic Distribution Power: What Are Tencent, ByteDance, and Alibaba Competing For?

In the race to dominate the AI era's entry point, China's tech giants—Tencent, ByteDance, and Alibaba—are aggressively deploying AI Agents to control the future of traffic distribution. Alibaba is pursuing a dual-track "closed loop + openness" strategy. Its Qianwen app is evolving into a super-Agent integrated across its ecosystem (Taobao, Alipay, etc.) to handle complex tasks like travel planning. Concurrently, it is opening its platform to external brands (Luckin Coffee, KFC) and has launched a B2B Agent platform, "Wukong," targeting enterprise automation. Its other flagship, Quark, aims to be an "AI super search box" for information and tasks. ByteDance is executing an omnipresent "sprawl strategy." Its Doubao app boasts over 300 million monthly active users and is evolving into a default AI entry point for daily life, with plans for paid versions and e-commerce integration. Its core weapon is the Kouzi platform, a visual "AI assembly factory" for developers to build custom Agents. ByteDance is also pushing hardware integration, collaborating on AI phones and developing smart glasses to embed Doubao everywhere. Tencent is playing its long-held "ultimate card" by quietly embedding an AI Agent directly into WeChat. This Agent, accessible via a swipe, can understand user commands and automatically execute tasks by calling upon WeChat's millions of mini-programs (e.g., finding and ordering coffee). This leverages WeChat's unparalleled 1.4-billion-user ecosystem to position the app as an AI-powered "service operating system," a move that could dramatically reshape the competitive landscape. The core battleground is shifting from competing for "user screen time" to competing to be the "default execution layer" for user intent. The business model is evolving from an "attention economy" to an "intent economy," where the Agent that can most efficiently fulfill a user's need gains control over service access and token flow. This represents a fundamental change in how users connect with digital services, making the fight for the Agent入口 (entry point) a pivotal moment for redefining industry leadership in the AI age.

marsbit1 ч. назад

Agents Take Over Traffic Distribution Power: What Are Tencent, ByteDance, and Alibaba Competing For?

marsbit1 ч. назад

From Banning Doubao to Embracing Honor: Why Did WeChat Suddenly 'Change Its Face'?

The article explores the sudden shift in WeChat's strategy towards AI assistants from mobile phone manufacturers, transitioning from strict opposition to active collaboration. For over a year, WeChat fiercely resisted attempts by phone AI assistants (like ByteDance's Doubao in late 2025) to control its features via GUI automation ("simulated clicking"), citing security and data control concerns. This stance created a significant barrier for system-level AI integration. Now, Tencent has initiated A2A (Agent-to-Agent) partnerships with major phone brands like Honor, Xiaomi, OPPO, and vivo. This model allows a phone's system AI (e.g., Honor's YOYO) to parse a user's voice command and send a structured request directly to WeChat's own internal AI agent via secure APIs. WeChat then executes the action (e.g., sending a message) and returns the result. The article attributes Tencent's "change of face" to strategic pressure. While leading in social app usage, Tencent trails rivals like ByteDance and Alibaba in standalone AI app popularity. WeChat, with its vast mini-program ecosystem, is Tencent's key asset for an AI comeback. The upcoming WeChat AI agent aims to handle tasks like booking and payments within the app. However, phone system assistants remain the primary AI entry point for most users. The A2A collaboration allows Tencent to extend WeChat's AI reach to this crucial system layer while maintaining control over its core functions and data. For phone manufacturers, embracing A2A is a pragmatic move. The GUI route proved unviable due to WeChat's blocks. A2A offers a compliant path to integrate a vital service, enhancing their AI assistants' usefulness. It allows them to focus on developing their own AI ecosystems for other services while cooperating on WeChat access. The collaboration is framed as a mutual, strategic necessity: Tencent gains a distribution channel, and manufacturers gain a key functionality. The partnership relies on a "dual authorization" mechanism for security, requiring both user and app consent for each action. While questions about long-term data privacy practices remain, experts note A2A is more secure and compliant than GUI automation. Ultimately, this cooperation is seen as a tentative, calculated truce. Tencent's long-term goal is to make WeChat an AI-powered "service OS." Phone manufacturers aim to make their system AI the central user interface. Their paths may converge or clash in the future, but for now, the A2A deal represents the opening chapter in the battle for the AI-era user入口, driven by necessity and strategic calculus on both sides.

marsbit3 ч. назад

From Banning Doubao to Embracing Honor: Why Did WeChat Suddenly 'Change Its Face'?

marsbit3 ч. назад

Торговля

Спот
Фьючерсы
活动图片