Vitalik: Reject the crypto doomsday scenario, be the last line of defense against AI tyranny

Vitalik Returns to Chiang Mai, Deeply Analyzes Prediction Markets, SocialFi, Decentralized Stablecoins, and Ethereum’s Role in the AI Era, Warning Crypto that Pure Speculation Will Lead to Death.
(Background recap: Vitalik lands in Taipei! Ethereum Foundation mission, next steps—full answers with EF new Executive Director)
(Additional context: Vitalik casually mentions “Ethereum’s Year of Great Leap”—L1 scaling will increase throughput tenfold)

Table of Contents

• Vitalik’s Chiang Mai Reflections: Mastery of Technology, Why Applications Go Astray
• Behind the $70,000 Profit: Vitalik’s “Counter-Crazy” Strategy and Oracle Concerns
• Survival Strategies for Ethereum in the AI Era
• Vitalik’s Conversation with 706 Community: About Motivation, Original Intent, and Ideals

A year apart, Vitalik seems to have undergone some subtle changes.

This was the first thought that popped into my mind after just finishing my second interview with him in Chiang Mai.

Rewinding to late 2024. Our first conversation took place in a quiet, enclosed indoor space on Nimman Road, Chiang Mai. At that time, he was full of excitement about Web3 application layer innovations—from Farcaster to Polymarket, Solana, Base—we talked for a solid 90 minutes.

This time, it’s late January 2026, and the scene shifts to an open space during Chiang Mai weekend.

That afternoon, Vitalik strolled from the “Four Seas Community” to the “706 Community Co-living Space.” On the second-floor balcony, he sat alone on a swing, swinging leisurely, with a relaxed demeanor that fit perfectly with any ordinary community member here. I sat down beside him, swinging gently, and threw a series of questions.

Around us, community members were busy with their own activities. No strict security, no pretentious display. Soon, a few curious 706 community members gathered around, sitting on the ground, resembling students chatting on a university lawn.

In subsequent conversations, I was surprised to find that his thinking system had undergone a substantial iteration over the past year.

Whether it’s Web3 social, prediction markets, or AI, his perspective has become sharper and more concrete. He dissected his observations on top projects like Polymarket, Farcaster, UMA, Chainlink, MetaDAO, and Base, and candidly shared his latest judgments on Ethereum’s role in the AI era, decentralized stablecoins, and RWA.

Of course, more unchanged aspects remain.

He still has no fixed residence, never staying in one city for more than two months; he still has no bodyguards, often lining up with us at the cafeteria for buffet; he still holds great enthusiasm for decentralized communities, tirelessly traveling between various Chiang Mai hubs.

When the conversation ends and night falls, the swing stops. As usual, he quickly runs into the street before it gets dark, hails a ride alone, and leaves.

Beyond this multi-billion-dollar decentralized kingdom, he continues to defend his freedom as an “ordinary person.”

Below is my latest dialogue with Vitalik. At the end, I also include selected questions from other 706 community members present.

Illustration: Vitalik at Chiang Mai 706 Community

Vitalik’s Chiang Mai Reflections: Mastery of Technology, Why Applications Go Astray

Joe Zhou: A year ago, on the eve of Devcon, I invited you for an interview titled “Vitalik, Forty-Two Days in Chiang Mai.” Today, a year later, we meet again here. What new feelings has this city brought to you personally during this return?

Vitalik: I’ve seen some communities thriving, like the Four Seas Community, which has undergone many different changes. There are lots of events, many people, and most importantly—they haven’t become boring.

Joe Zhou: Environment and time often reshape thoughts. A year later, I’m curious—how have your thoughts on core crypto issues evolved? Where are your current focus points?

Vitalik: The biggest change is that I see a huge disconnect between technology and applications.

Over the past year, Ethereum has made great progress in scaling technology. Our Gas limit has risen from 30 million to 60 million, aiming for 300 million this year. Successful implementations like zkEVM, and significant improvements in wallet infrastructure and user experience—technologically, it’s been very successful.

But on the application side, I see many concerns. Reflecting on five or ten years ago, the community had very diverse and grand visions for the ecosystem. Back then, everyone was hopeful—wanting to build DAOs, decentralized apps that truly change social cooperation, like a “decentralized Uber.” However, I feel many have since forgotten these original intentions.

Crypto succeeded in finance but lost its way in governance. For example, the current DAO “token voting” mechanism has flaws. The recent explosion of memecoins, like Trump’s Meme coin in early 2025, is a typical example. When he greedily issued a second token, MELANIA, his first coin TRUMP was already effectively dead.

Joe Zhou: Last year, we discussed SocialFi apps like Farcaster in depth. A year later, from today’s perspective, how do you evaluate their development?

Vitalik: SocialFi is in a somewhat awkward stage now. Its biggest structural dilemma: if you tie social and finance too tightly, financial incentives tend to undermine social incentives.

When users no longer seek quality content but come mainly to make money, they start creating大量垃圾信息（大量垃圾信息）。This is a dangerous signal—because financial incentives are destroying the essence of social interaction.

I like Substack’s model. If you look at the top ten authors on Substack, they are all thoughtful and substantive. But on some top SocialFi platforms in crypto, the top ten are often just click farms or hype accounts. The difference is: Substack curates and builds community. They work hard to find quality authors and help bring them onto their platform, not just provide a token issuance tool. This is something crypto entrepreneurs need to learn.

Joe Zhou: That explains Farcaster’s recent shift—why they no longer focus solely on social but pivot to wallets?

Vitalik: Yes. They couldn’t find a way to scale bigger. They’re not satisfied with a “small but beautiful” product; they aspire to reach tens of millions or even hundreds of millions of users. Under their current trajectory, they believe wallets are more likely to achieve this scale of mass adoption than pure social.

Joe Zhou: A few years ago, there was a common industry consensus that application layer would soon see a “big explosion,” but that didn’t happen. Looking back four years, did you share the same optimistic expectation back then?

Vitalik: Yes, I did. My thinking was: applications haven’t taken off because the core bottleneck is the underlying technology—like insufficient scalability, slow speeds, poor user experience.

But by 2025, at least on Layer 2 (L2), these technical barriers had been largely addressed. Yet, it’s awkward that we still haven’t seen a large wave of good applications. The only sector that saw a real explosion in 2025 was prediction markets, but even they revealed significant issues.

Joe Zhou: What specific issues do you see?

Vitalik: If you look at discussions on Twitter, Polymarket’s most promoted topics are “Which team will win next week” or “Bitcoin’s price in an hour—up or down.” I think, in the long run, these short-term bets don’t have much social significance. Prediction markets are successful as a tool because they work, but we need more meaningful applications.

Mechanisms with long-term incentives would be better. For example, Robin Hanson’s Futarchy—prediction market governance—is very interesting. In traditional governance, people vote for leaders (presidents, legislators) or for policies (“Should we build this road?”). Robin Hanson’s idea is: people vote only on “goals” (e.g., GDP growth, unemployment reduction), and then use prediction markets to decide “means.” Traders, motivated by profit, bring out the most truthful data. Currently, MetaDAO is experimenting with this approach.

Behind the $70,000 Profit: Vitalik’s “Counter-Crazy” Strategy and Oracle Concerns

Joe Zhou: Do you still use Polymarket? I remember you used it quite frequently last year.

Vitalik: Yes, I made $70,000 on Polymarket last year.

Joe Zhou: How much capital did you start with?

Vitalik: $440,000.

Joe Zhou: Many people lost money—how did you make yours?

Vitalik: My method is simple: I look for markets that are in “crazy mode,” then bet that “crazy things won’t happen.” For example, a market betting “Will Trump win a Nobel Peace Prize?” or during extreme panic, predicting the dollar will go to zero next year. When market sentiment enters this irrational “crazy mode,” I bet against it, and it usually pays off.

Joe Zhou: What specific sectors on Polymarket do you focus on? Crypto? Politics? Entertainment? Economics?

Vitalik: Politics and tech. If you want to make money, you should look at markets where people are making wildly irrational predictions—you can profit from that.

Joe Zhou: As the founder of Ethereum, do you have insider information? For example, when Venezuela is at war, some netizens seem to know insider info in advance. Have you encountered similar situations?

Vitalik: I want to highlight a case about Oracle vulnerabilities. There was a prediction market about Ukraine’s battlefield—whether Russian troops would control a certain city. The contract’s “control” standard was whether they controlled the key train station, with data sourced from ISW’s Twitter and maps.

One day, ISW’s staff—perhaps by mistake or intentionally—blacklisted their own system, and their map suddenly showed Russian control of the station. This caused a previously unlikely event (5% chance) to instantly become 100% in the prediction market. Although ISW retracted the update the next day, the payout might have already been made.

This reveals a huge problem: current Oracle data sources (like Web2 news sites, Twitter) have very low security standards. They never imagined that a piece of info they publish could determine the ownership of millions of dollars on-chain.

Joe Zhou: That does sound crazy. You just pointed out Oracle issues—how can we fix this?

Vitalik: Currently, there are two main paths to solving Oracle problems.

One is centralized—trust a company like Bloomberg to provide accurate info.

The other is Token Voting—decentralized. The idea is: let token holders vote on “what is true.” UMA exemplifies this model. (Note: UMA is a decentralized oracle protocol on Ethereum, relying on token holder votes to arbitrate data truthfulness)

Recently, trust in UMA has declined. People see a game-theoretic flaw: if whales coordinate to manipulate votes, ordinary users can’t fight back. Because in this mechanism, even if you’re right, if you oppose the majority, the system will declare you wrong, and you’ll lose money. This forces users to follow whales’ votes rather than the truth.

I believe a trustworthy Oracle is very important. Because all DeFi relies on Oracle data. If you want to build real-world applications—like tokenizing real estate or predicting elections—you need an Oracle. Currently, most in DeFi trust Chainlink, but its mechanism is complex and somewhat centralized.

I hope we can find a better solution in the future.

Illustration: Vitalik sharing a book club session at Chiang Mai’s Four Seas Community

Survival Strategies for Ethereum in the AI Era

Joe Zhou: I want to discuss a hot topic: AI. Over the past year, the market has placed enormous hope on the integration of AI and crypto, but now the atmosphere seems to have shifted into collective confusion. In this new AI era, what role do you think Ethereum should play?

Vitalik: Returning to fundamentals, Ethereum is a decentralized world computer (World Computer). Its core attribute is “permissionless”—whether humans, companies, or AI agents, all have equal access. This means AI can hold assets, trade, and even participate in DAO governance on Ethereum. From this perspective, Ethereum is already prepared.

Joe Zhou: But the confusion lies in: what exactly is the integration point? How do we concretely realize these grand concepts?

Vitalik: First, we must beware of a thinking trap: don’t integrate just for the sake of integration.

For example, even in the AI era, the underlying TCP/IP protocol (Internet protocol) doesn’t need to be overhauled because of AI. Similarly, blockchain as a trust protocol may not require a radical change.

However, if we look at the application layer for intersections of AI and crypto, I see several promising directions:

1: AI’s bank accounts. AI can’t open accounts at traditional banks. If an AI agent needs funds to perform tasks, crypto is its only option.

2: Prediction markets. AI can act as a trader, providing more accurate information.

3: Content authenticity. Using blockchain to prove whether content is human-created or AI-generated.

Joe Zhou: There’s a popular term now—“Vibe Coding” (coding easily with AI assistance, not focusing on details). Do you still code manually sometimes?

Vitalik: Sometimes. I still keep the habit of coding myself. My coding work mainly falls into two categories:

First, practical scripts—small programs for personal use, mainly to improve my productivity.

Second, research validation. When exploring complex cryptographic algorithms, I write implementations (usually in Python) to verify my mathematical ideas.

Joe Zhou: Do you use popular AI coding tools like Claude, Gemini, or Manus? Which one do you prefer?

Vitalik: I don’t rely on any specific tool. I mainly use OpenRouter, an aggregator platform that allows me to call various models. For coding, I still use mainstream tools like ChatGPT, DeepSeek, Gemini.

Vitalik’s Conversation with 706 Community: About Motivation, Original Intent, and Ideals

(The following is a joint interview compiled by Joe Zhou and 706 community members with Vitalik)

706 Community: What is your current motivation for doing things?

Vitalik: My motivation mainly comes from three levels—or three senses of urgency.

First, to prevent the “Crypto doomsday scenario.” I’m most worried about a future where the entire industry degenerates into a 100% speculation zone—only trading, no real applications. If that happens, once everyone gets bored, the industry will die in boredom. To avoid this, we must build real value—better DAOs, truly integrated decentralized applications across industries, and more open DeFi.

Second, to make Ethereum’s technology better. Honestly, Ethereum’s current tech isn’t good enough. Although Layer 2 solutions have addressed scalability, they are still mostly highly centralized. We need to push Layer 2 to be more scalable and truly decentralized, so that user experience can match Web2.

Third, if we fail in crypto, the future tech world might be dominated by centralized AI, which would be very dangerous. Crypto is our line of defense against this digital authoritarian trend, maintaining diversity and freedom in the tech world.

706 Community: Hypothetically, if you could discard all of Ethereum’s historical baggage and redesign it from scratch, how would you do it?

Vitalik: Honestly, the technical roadmap wouldn’t change much. Because no matter how many times you start over, Ethereum’s core goal remains the same: to be a decentralized application platform.

Looking back before blockchain, the most successful decentralized network was BitTorrent. It’s great, but it lacked a key component: a global shared state. With BitTorrent, you can share files, but you can’t create a currency, assign ownership, or run a DAO. Without “state,” you can’t record “who owns what.”

So, if I were to redesign, to fill this gap, the platform must have two core features: first, scalability. Without high scalability, on-chain costs are too high, limiting applications to high-net-worth DeFi trades; mass adoption becomes impossible. Second, speed. Only fast enough can provide a usable user experience.

706 Community: As an ETH holder, I want to ask a sharp question: what do you think is the biggest yet most overlooked risk Ethereum currently faces?

Vitalik: Honestly, I’m less worried about technical risks now.

What truly concerns me is the application layer. The “failure scenario” isn’t network crashes or hacking; it’s that thousands of applications are developed, but upon review, none are truly socially meaningful. If we have the strongest decentralized tech but only build toys or gambling platforms, that’s the biggest risk.

706 Community: Looking ahead 5 to 10 years, what do you see Ethereum’s role as?

Vitalik: I hope it becomes a core hub for all decentralized applications—not just finance, but penetrating all industries.

Its core value is “true ownership.” If you own something, it genuinely belongs to you. In the traditional world, you’re always controlled by big companies—they can censor you, change rules, or impose fees. We need to change that. This isn’t just about finance; it’s also crucial in social and identity domains.

So, Ethereum’s success depends on our technology (scalability, experience) being strong enough to support these applications and serve as their solid foundation.

706 Community: Will developments in AI and quantum computing pose a 51% attack risk to Ethereum?

Vitalik: I don’t think so. We need to clarify: a 51% attack targets the consensus and coordination mechanism of PoS (Proof of Stake)—which requires controlling 51% of the total stake, not hash power.

Quantum computing mainly threatens cryptographic signatures, not consensus mechanisms; and I see AI as more of an aid than a threat. For example, AI can help us with formal verification, discovering vulnerabilities, making Ethereum safer.

706 Community Members: Are you paying attention to Hyperliquid?

Vitalik: Not really, not much.

706 Community: You mentioned wanting more applications—what types of applications do you most want developers to build?

Vitalik: First, decentralized social (DeSoc). Current social media has huge issues. Although most people dislike X (Twitter), we lack a good alternative. Users are locked into platforms, with no freedom to “move.” We need to build a truly user-owned, portable social network.

Second, “smarter” DAOs. DAO remains a valuable concept, but it needs to be smarter. Not just issuing tokens and voting. Developers should think more deeply: what are the specific goals of this organization? What governance structure best fits these goals? We need more experiments—try different approaches.

Third, more decentralized stablecoins would be better.

706 Community: Decentralized stablecoins—are you referring to those pegged to fiat currencies?

Vitalik: The most interesting part is—if we can create “non-fiat pegged” stablecoins, that would be real innovation.

Joe Zhou: If not pegged to fiat, what would they be pegged to?

Vitalik: To real-world values. For example, pegged to CPI (inflation index), so your money always equals a certain amount of bread; or pegged to energy prices. That’s true “stability.”

Joe Zhou: That logic sounds very much like Facebook’s Libra project—a basket of currencies.

Vitalik: Exactly. I think Libra’s idea was good, but the execution went off track. They turned a decentralized currency vision into a Zuckerberg-controlled corporate version. Due to Facebook’s poor privacy record, people naturally fear it. So we need to build a similar but truly decentralized version.

Joe Zhou: Do you do any interesting “personal experiments” in life or tech recently?

Vitalik: (pauses to think) I’m trying to completely detach from the official X client. Now I mainly use decentralized aggregators like Firefly to publish and browse content.

706 Community: What’s your ideal Web3 social product?

Vitalik: It doesn’t have to have features that Twitter (X) doesn’t have; nor necessarily new financial features. But it must have higher quality. The most important isn’t features but who the users are.

706 Community: Most ordinary users don’t care about “data sovereignty,” which might be why Web3 social hasn’t taken off. At this stage, where do you see the breakthrough?

Vitalik: Honestly, I don’t know yet.

Note: Thanks to Qiuxiu and other 706 community members for their help with this article.