Guests: Arthur Hayes, CIO of Maelstrom; Illia Polosukhin, Co-Founder of NEAR

Moderators: Andy; Rob

Podcast Source: The Rollup

Original Title: Arthur Hayes & Illia Polosukhin: Privacy Is The Last 1000x (NEAR & ZEC)

Release Date: May 25, 2026

Key Takeaways

This episode of The Rollup features Arthur Hayes and Illia Polosukhin discussing macro liquidity, privacy assets, NEAR Intents, AI and on-chain execution layers, and investment theses around HYPE, NEAR, and ZEC. Arthur argues that war, AI arms race, and supply chain reconfiguration are driving the US, China, and Europe to continue supporting their economies through debt and monetary expansion, with liquidity eventually spilling into Bitcoin and a select few crypto assets with real narratives and revenue. Illia emphasizes that for blockchain to enter everyday payments, salaries, invoices, and AI agent economies, privacy is not optional—it’s a prerequisite for mass adoption. Both agree the crypto market is shifting from indiscriminate speculation to fundamental screening, where privacy, sovereignty, real revenue, and token value capture will define the next phase.

Highlights Summary

Macro Liquidity & AI Arms Race

L1 Blockchain Convergence & Crypto Fundamentals Resurgence

Zcash & Privacy Assets Undervalued

NEAR Intents, Private Transactions & Mass Adoption

AI Agents & On-Chain Execution Layer

Hyperliquid & the Dream of Decentralized Finance

AI Labor Substitution & Political Risk

Arthur's Macro Investment Thesis

Moderator Rob: Arthur, you were once cautious, warning about risks, then turned fully bullish. What changed? What triggered the shift?

Arthur Hayes:

At the start of the year, I wrote an article arguing that Bitcoin was already pricing in a credit event driven by AI-induced deflation. My thesis was that monetary authorities wouldn’t print enough money until a financial crisis was visible. At the time, Nasdaq had flatlined since Bitcoin hit its all-time high, while Bitcoin dropped alongside U.S. investment-grade bond ETF IGB—from $126k down to ~$60k—indicating a clear credit event.

Then on February 28, the U.S. opened fire on Iran, igniting a full-scale conflict. Markets realized this would be a liquidity-friendly scenario. First, AI is now embedded in national defense: drones, AI-driven UAVs, and intelligence systems are actively involved in warfare across all fronts. Governments will print money to win wars—and since AI is part of the war effort, both the U.S. and Chinese governments will underwrite AI capital expenditures.

We’re already seeing bank loans and equity flows toward semiconductor manufacturers. In the U.S., Intel has received support, and similar announcements exist in quantum computing, such as a $1B investment into IBM-related projects. These are all part of the same theme.

Meanwhile, many nations previously assumed they could rely on contested waterways like the Strait of Hormuz for food and energy. Now that’s becoming untenable. So they ask: Why hold U.S. Treasuries? If I need to feed my population or operate on a remote island without aviation fuel, holding Treasuries won’t help—especially if my ships can’t even pass through the strait.

Thus, they must build redundant supplies for critical goods—especially food and energy; invest in new trade relationships; construct pipelines to reroute oil from the UAE instead of the Persian Gulf. All of this means savings held in U.S. Treasuries must be sold and converted into tangible goods.

The U.S. won’t allow such a sell-off to crash the market. It will print money to fill the gap, ensuring systemic stability. That’s why I view February 28 as the catalyst. The U.S., China, and Europe will all print money to fund wartime economies and AI capital spending—money that will ultimately flow into Bitcoin, explaining its strong performance.

Heavy Bets on NEAR, ZEC, and HYPE

Arthur Hayes:

We’re now seeing a small group of tokens rally. Among them, NEAR, HYPE, and Zcash—assets I hold—have performed exceptionally well since February 28. This is my core logic.

I didn’t do much trading in Q1. After the war began, I added some positions, but fundamentally, we’ve long held these assets at very favorable prices. Now the market is validating why we bought them 6 to 12 months ago.

Moderator Andy: The entire crypto Twitter is debating whether this trade has become consensus—but we’ve already gone all-in. We’ve been focused on macro trends and followed your CoinDesk interviews and articles. Initially, the market was uncertain, but then NEAR, HYPE, and ZEC suddenly surged—so everything is now coalescing in this integration phase.

Layer-1 Convergence Has Begun

Moderator Andy: Illia, I’d love your take on market convergence. This isn’t your first bear market—we’ve seen these phases before—but this one feels especially real because the ecosystem itself is evolving.

Institutional allocators have entered, macro conditions are unstable, pushing people to adopt an “adapt to institutional era or get left behind” mindset. Meanwhile, the industry is maturing—more complex, requiring products, revenue, real users, and truly sustainable, robust tokenomics.

From a founder’s perspective, what happened over the past 6–8 months? Why the rapid change? How did NEAR navigate this phase?

Illia Polosukhin:

I see several converging threads. Start with Layer-1 and Layer-2 logic. I recall a talk I gave in 2019: we’d go through a massive boom, then enter a consolidation phase, with only a few projects surviving. We’re approaching that stage now.

Blockspace is now a highly commoditized good, with supply vastly exceeding demand. The historical issue was its lack of interoperability. NEAR’s bet on chain abstraction and Intents is making blockspace fungible. Meaning, every chain, every asset, every user can truly interconnect—without needing to think about which specific blockspace they’re using.

Second thread: fundamentals returning. You’ve been preaching this for over a year—maybe even years. It’s finally happening. Whether it’s the market or institutional investors analyzing markets—not BlackRock or Fidelity, but actual funds and professional investors—they’re shifting.

The old logic is fading: buy an asset because retail will keep buying it. Retail risk appetite has dropped sharply. As Arthur just said, people care more about whether they can afford oil and food next year than whether to speculate on an asset.

So the market now asks: Which assets generate real revenue? Which provide products we actually use? If I stake it, can I gain new capabilities? For example, staking HYPE grants lower fees and greater market access. ZEC offers privacy. NEAR provides cross-chain, intent-based, and computational capabilities.

These are exactly the new capabilities market participants want. Thus, these assets are becoming focal points—not those that may be held, but whose value is unclear until fee switches open. I believe these two threads are converging, shaping today’s broader crypto environment.

Why Arthur Loves Zcash

Moderator Rob: The industry seems to be entering a restructuring phase. Bankless folks appear to be splitting—David Hoffman sold all his ETH, Merch no longer supports Solana, fully pivoting to Zcash. Arthur, why is Zcash and privacy so important to you? Is this the core ethos of crypto? And what role does NEAR Intent play here?

Arthur Hayes:

Zcash’s current momentum is certainly new, but I remember how hot it was in 2016 before mainnet launch. Back then, I was at BitMEX—we launched the first Zcash price derivative. The market was insane; prices on Poloniex once hit ~$3,000. If you recall that exchange, there was even a moment when 7 BTC exchanged for 1 ZEC.

That was chaotic. Then due to block rewards, supply exploded, and prices collapsed. But Zcash had issues back then—like trusted setup, requiring trust in Zooko, Eli, and others during key generation rituals. There was also controversy around the 20% block reward subsidy.

Those problems are now resolved. The protocol has upgraded, trusted setup removed, and we know it’s cryptographically secure. The 20% subsidy is gone. Now it has a clean slate, sufficient supply to form a real market—not the extreme volatility of genesis.

Naval Shifted Arthur’s Perspective

Arthur Hayes:

Last year at Token 2049, I attended a dinner where Naval was present. He started talking to me about Zcash. At the time, Zcash had risen from ~$30 to ~$120. I asked him why he liked it, knowing I held significant Monero.

He told me Monero’s ring signatures aren’t as strong as commonly believed. Japanese law enforcement had successfully de-anonymized transactions in a criminal case. After hearing that, I thought: this guy is credible. So I went home and bought millions in Zcash.

I also noticed something: I have many brokers, but only two would quote me Zcash prices—because many institutions say, “We don’t trade privacy coins.” My reaction? If I can’t easily buy it, I want it even more.

Then I dug deeper, validated Naval’s claims. I kept buying, and kept adding. Zcash boomed in 2016 because people knew Bitcoin is pseudonymous, not private. Many genuinely value privacy in certain contexts. Having a private currency on the internet is perfectly valid—this is what Zcash represents, and what Monero represents.

As we increasingly see that Big Tech, Big Government, and AI can know everything about us, track our lives, cryptographic proof of monetary privacy becomes essential. That’s why I made Zcash my second-largest holding. Even though I bought after a 4x rebound from $30, the performance remains stellar.

Next challenge: UX. The best apps used to have other names—now called Zashi. Users can now hold shielded Zcash there. In 2016, one major problem was that while Zcash worked, almost no one could hold it shielded. Everyone used transparent Zcash—effectively a subpar version of Bitcoin. If you can’t hold it privately, what’s the point?

Now it’s different. Apps like Zashi let you hold shielded Zcash on mobile with great UX. Pair it with cold wallets like Keystone, and you can store shielded Zcash offline. I strongly urge all Zcash holders to hold it in shielded form—otherwise, why hold it at all?

NEAR Intents & Anonymous Exchange

Arthur Hayes:

UX is now excellent. Next comes the exciting part. We invested in Zashi’s funding round. Within the Zashi app, I can send any crypto asset anonymously to anyone online via shielded Zcash and NEAR Intents. For example, I start with shielded Zcash, end up with USDT on Tron, and send it anonymously.

This is huge. Look at NEAR’s price—like other crypto assets, it’s far below its all-time high and has cycled through downturns. I thought: if privacy becomes a dominant narrative, Zcash is the first stop. The second is whoever enables anonymous value transfer across chains. That capability is equally powerful.

NEAR’s economics will catch up—this is a highly asymmetric opportunity. Of course, I haven’t allocated as much to NEAR as to Zcash, but I believe NEAR has 20x potential in the next year, while Zcash may be 5x. You allocate capital based on risk. That’s why these two assets form the core of my privacy thesis: in a world of AI, Big Tech, and Big Government, privacy will be rediscovered by the market—and Maelstrom will benefit.

Privacy Is a Prerequisite for Mass Adoption

Moderator Rob: Illia, when you look at NEAR from Arthur’s capital allocator lens, clearly AI is a central theme—supporting identity management, payment networks, and infrastructure for AI agents. But the true driver unlocking Zcash’s ecosystem value lies in Intents, which make privacy coins more efficient and practical.

Here’s some data: NEAR Intents’ historical transaction volume is nearly $20B, currently ~$18.9B. From these, NEAR Intents has generated $33M in fees, with a large portion coming from Zcash—since this is currently the only place capable of executing such operations. Arthur also mentioned privacy intents create positive protocol cashflow.

How do you plan to monetize Intents? What’s NEAR’s current cashflow situation? And how do you plan to expand NEAR Intents to better serve the entire NEAR ecosystem?

Illia Polosukhin:

Arthur’s core point: we need on-chain privacy. I’d go further: if we want blockchain to enter daily life, privacy is non-negotiable. It’s actually the precondition for widespread crypto adoption.

If I buy coffee at a café, I don’t want the café to know my balance—or the world to know I just paid there. I don’t want someone waiting for me next time I visit. Recently, someone on X discussed a person’s on-chain credit card transaction—everyone could see it. That’s absurd.

So, if we want crypto, stablecoins, and other assets to become real-world currencies, privacy is foundational. Even for investing, I now need multiple addresses—one for payments, one for receipts—each isolated, managed in spreadsheets. That’s unacceptable in normal life.

Privacy Intents aim to deliver not just sovereign, censorship-resistant, private assets like Zcash. Zcash is vital—we absolutely need it. But we also want to solve another problem: how to confidentially transfer, trade, pay, earn yield, and execute other actions across 150+ supported assets.

Privacy Intents essentially build a private shard atop NEAR. Users enter this private shard, conduct internal transactions invisible to outsiders. No complex cryptography required on the client side—all math happens internally. It’s lightweight, programmable. You can write and deploy smart contracts inside.

This means we can already support transactions, transfers, and private payments. More features will roll out via partnerships and ecosystem integration. I’m excited because this is the execution layer of the privacy thesis—making privacy scalable, accessible, and usable.

Take salaries: no one uses crypto to pay wages because everyone sees who earns what. Or invoices—many use cases we’ve long claimed crypto should enable. When everything is transparent, they simply don’t work. This is incredibly exciting—it will boost transaction volume and count.

We charge fees on every transaction and use them to buy back NEAR. The model is direct. Our goal is to keep doing this. Meanwhile, we’ve already reduced inflation across the ecosystem. In November last year, we halved inflation. NEAR is now fully diluted, and I’ll continue pushing the community to reduce inflation further. As ecosystem revenue grows, we can balance the ledger, turn NEAR deflationary, and achieve true economic sustainability and profitability.

Arthur Hayes:

I’ll add to Illia’s point: full dilution is crucial, especially when evaluating whether a Layer-1 can truly perform in an uptrend. Many Layer-1s do flashy things, but still carry massive unlock risks from venture capital firms ready to dump.

You want clean upside. NEAR achieves this simply because it’s existed long enough to complete the cleanup process. It’s a strong asset—open sky ahead, no one poised above me to flood the market with unlocks.

AI Blockchain Vision Since 2017

Moderator Rob: NEAR is among the few fully diluted tokens, especially within L1 blockchains. It already generates over $30M in cashflow, with Intents being a primary revenue driver. Illia, can you elaborate on NEAR’s other AI features—NEAR AI, IronClaw—and your current product suite? What excites you most about the product stack? How does it integrate with Privacy Intents? And what’s the revenue potential, and how will it impact NEAR’s overall economy?

Illia Polosukhin:

The top-level vision: in the future, we’ll use AI for computation, and blockchain will be the execution layer for everything. So we’re building both parts and connecting them deeply.

AI also needs privacy. You don’t want labs harvesting your data to train better models, then selling services back to you via subscriptions. You want richer context, more access—but if the model runs on a third party you don’t trust, it’s dangerous. For example, giving away encrypted account or bank access permissions could lead to data harvesting or unauthorized transactions—without your knowledge.

What we’re building on the AI side is an agent experience rooted in privacy and security. It’s essentially an execution arm: today you manually handle payments, but tomorrow, your agent handles everything—from daily shopping to hedging spot positions on Hyperliquid to stacking prediction markets.

For example, you say: “I want to open a position. I believe something will happen tomorrow—help me build the right strategy.” It finds the right assets, combinations, and sets up the position. Today, doing this manually would be extremely tedious.

IronClaw already has a working example. Input your crypto address, and it analyzes all your assets and positions, suggesting better yield opportunities and optimal reconfigurations across protocols. This can be massively expanded.

So these two layers work together. AI is the new compute interface; Intents are the underlying business layer. It’s not just for payments—it extends to supply chains, commodities, and broader transactions. All will flow via Intents, because AI finds, negotiates, and settles deals far better than traditional email, invoices, and billing systems.

We have a concrete example: the Agent Marketplace. Agents can hire other agents to complete tasks or deliver goods. This is a glimpse of the future—even early, real-world enterprises are using it. You can hire an agent to procure supply chain parts, build a marketing website, develop an app, or draft investment copy.

These agents run on our verifiable computation. You know what they’re doing—you can audit it. They run on our secure infrastructure, so you can grant them access to contacts or internal documents. Historically, hiring employees meant giving them access. But historically, granting third-party access was hard. Now, if the agent runs in a verifiable, secure environment, you can safely share company data.

What I’m describing is changing how labor and supply chains work. So our target market is the total serviceable market for all labor and supply chains—running entirely on Intents.

Moderator Andy: You can see the convergence between AI and Intents: agents use Intents to transfer value, and users want to use private Intents to transact with Zcash or move value. Some might say NEAR’s direction is scattered—but from the beginning, it seems you’ve had a clear thread.

Illia Polosukhin:

Yes. We started in 2017 with the core belief that AI would become how we build software and interact with computation. That was 2017. Later we realized to truly achieve this, we needed blockchain. So in 2018, we focused on NEAR Protocol. These elements have always been part of NEAR’s DNA.

Hyperliquid Realizes the DeFi Dream

Moderator Andy: Arthur, I know everyone wants to hear your take on Hyperliquid. HYPE is now around $61, and privacy transactions represented by Zcash and NEAR differ significantly from Hyperliquid’s logic—but they feel like pieces of the same puzzle.

Privacy is crucial—core to the industry—it empowers freedom. But in today’s world, especially in finance, privacy is severely undervalued. On the other hand, Hyperliquid represents the best chance for DeFi and decentralized exchanges to surpass centralized finance. Many have forgotten: we entered this space partly to replace traditional finance, creating a real alternative system.

That dream was consumed by inflated valuations, low liquidity, scam projects, and infrastructure narratives. But Hyperliquid appears to be reviving that dream. Though these investment theses differ, they attract the same crowd. What’s your take on Hyperliquid? Where is it headed? Why are you so bullish?

Arthur Hayes:

One of crypto’s killer apps? Exchanges. Who are the wealthiest people in crypto? Exchange founders. BNB remains among the top 5 largest crypto assets—yet it’s not truly a cryptocurrency; it’s CZ’s server and the chain he launched.

We know how to make money in crypto. Many just complicate things—infrastructure, real-world assets, etc. The truth: exchanges make money. Clear objective: we have the internet, blockchain—let anyone anywhere trade anything, add leverage to make it fun.

I’ve been in centralized exchanges and always knew the trend was toward decentralized ones. Early pioneer: dYdX—the first star in decentralized exchanges, with stellar performance in 2020–2021. But it drifted off course. dYdX wanted to do what Hyperliquid does—just Hyperliquid executes better, while dYdX’s tokenomics were flawed.

Then came GMX in 2023’s downturn. Solid model, but room for improvement in tokenomics and listed assets. Enter Jeff and his team—coming from high-frequency trading backgrounds, brilliant engineers who delivered excellent code. Most importantly, they fixed the tokenomics. Perpetuals aren’t new—we created them in 2016. DEXs aren’t new—existing since 2018–2020. The key was getting tokenomics right.

Hyperliquid has no VC sales—only team allocation—and nearly all revenue flows back to token holders. No other project matches HYPE’s revenue scale in returning value. That’s why it’s so successful—and why users are deeply engaged.

When they launched HIP-3, enabling permissionless market listings, users could trade Nasdaq, S&P, oil, etc. In December 2023 and January 2024, these were tiny markets. But because politicians love stirring trouble on weekends, traditional markets lacked price discovery—giving them a three-day window. Now Hyperliquid is the sole price discovery venue on weekends—enough liquidity, accessible to all.

This isn’t just for those with U.S. brokerage accounts. Traditional financial media now cover Hyperliquid—they have nowhere else to go. If they could discuss oil futures on Saturday at CME, they’d do it—and ignore Hyperliquid completely. But they can’t. So they can’t ignore Hyperliquid—it’s the only place to trade. Everyone accesses data, everyone trades.

This creates a flywheel. More people learn about Hyperliquid, understand revenue returns to token holders; stake enough HYPE for fee discounts; participate in listing your own markets—turning it into a self-fulfilling prophecy.

I believe it’s already broken its all-time high and will go much higher. Its easiest entry point is capturing centralized exchange trading volume. Hyperliquid is now maybe 7–8% of that—future growth is inevitable as it lists more assets, offers higher leverage, and improves usability.

Even capturing 10–15% of Binance’s perpetual volume would push HYPE far beyond current levels. It doesn’t need reinvention—just capturing existing traders. These users already hold stablecoins or other crypto, trade on centralized exchanges, pay high fees, want a better experience—and want real ownership of an exchange.

AI Labor Substitution Investment Thesis

Moderator Rob: Illia, you mentioned the total serviceable market for human labor. To help viewers grasp this concept: global wages and compensation total ~$11.7T annually. Meanwhile, AI agents are gradually becoming key players in the labor market.

Arthur, your AI labor substitution thesis suggests this will collapse consumer credit. There will be bumps, but ultimately trigger massive money printing. I’m curious: how do you see AI replacing labor? Will it evolve into agents? Will these agents run on-chain? How will the economy respond? Arthur first, then Illia—how should we prepare for this inevitable shift to ensure a smooth transition?

Arthur Hayes:

AI’s labor impact depends heavily on where you live. Let’s start with the U.S.—most viewers likely come from or live there. The unemployed Americans now are the most protected: coastal, high-income, white-collar, college-educated professionals.

When China’s low-cost labor joined the global workforce, Rust Belt blue-collar workers lost jobs—but no serious discussion occurred. No one seriously debated it in 2005 when the Rust Belt was hollowed out. Later, Trump and others gave them a voice—but only after they’d already lost their jobs.

Now it’s different. In the U.S., it’s educated, coastal, high-income, politically active people losing jobs. They’ll be protected—I don’t know the exact solution, but they will be.

But workers in India, Bangladesh, Philippines doing backend work for U.S. companies? They’ll lose jobs instantly—no one cares. They may end up homeless, starving, facing social chaos. Real violence may erupt there. You won’t see it in mainstream news—but that’s the difference.

They can’t fix it with money printing. They’ll suffer greatly. The U.S. can mitigate via printing because the dollar remains a reserve currency. This affects financial markets, which mainly care about Western high-net-worth individuals. Sadly, regional divergence will be extreme.

Illia Polosukhin:

I believe replacement has already begun—though unevenly. The only real chance is to stay at the forefront. We’ve always said: use these technologies to enhance your ability, learn faster, apply faster.

Huge opportunities exist. If you don’t know something, AI knows—and can guide you. If you’re outside a country, you can start a business accepting global currencies via Intents and broader crypto systems. This is arguably the easiest era ever to start and globalize a business.

But the condition: you must act. You need initiative and sovereignty to execute. That’s why the Zcash, HYPE, NEAR thesis resonates: people want to reclaim agency and act—returning to crypto’s original core.

So I believe this is the time to commit—if you’re willing to seize the opportunity.

Moderator Rob: Arthur, from a capital allocation angle: is the logic that we’ll see consumer credit pressure, a bumpy middle phase—so not a broad market rally? But that explains why we’re seeing a more structural bull market. Illia’s sovereignty thesis and privacy narrative are responses to a larger story.

Arthur Hayes:

Liquidity won’t flow everywhere. Bitcoin’s advantage: if tomorrow’s fiat unit increases in supply, its price mathematically rises—pure math. Everything else relies heavily on narrative.

Look at global capital markets: only AI stocks are performing well; most others are terrible. Because most people are negatively affected by AI. Maybe you sell products to the unemployed, customers stop spending, or certain labor disappears. Or you’re squeezed out of credit markets as AI giants absorb all available credit.

Every government now believes they must fund the AI capital boom—pushing other priorities aside. So while we all marvel at this incredible bull market, with some crypto assets doing well—really, the core driver is AI stocks. I expect them to keep rising—but this isn’t a universal tide lifting all boats.

Political Risk Scenario

Moderator Rob: You’ve touched on similar views in recent articles. When do you think this “music” will stop? If it does, what signals should we watch? When will we realize AI capital spending-driven money printing might slow?

Arthur Hayes:

I think it might be a super-large IPO—but I don’t think SpaceX’s size is enough. They’re reportedly raising ~$75B, but I haven’t deep-dived into details.

But I’m increasingly convinced it’ll be political. In the U.S. and Western economies—where capital is concentrated—a message is emerging—and I believe it’s highly compelling: human-to-human interaction generates data, which made Elon and others extremely wealthy. Where’s our share?

AI can’t exist without the sum of 10,000 years of human knowledge and interaction. But what do I or my community get? Higher electricity bills, water pollution, mineral resources diverted for data centers. What do we get?

So I believe a logical political message will emerge in the U.S.—one that attracts both Democratic and Republican voters. I suspect AOC could be the 2028 Democratic presidential candidate. Investors would panic.

I disagree with much she says—but if you weren’t a holder of assets up 50x in the past year, her arguments make sense. Investors will realize: if AI profits face massive progressive taxation, would you still pay 100x earnings for a company potentially taxed at 50% on AI gains? Absolutely not.

If she starts polling near Rubio, J.D. Vance, or any Republican successor to Trump, you’ll start thinking: maybe I should cash out part of my position. Once you start questioning the future, before election results are clear, the market ends.

So while a mega-IPO could spark a reversal, I’m increasingly wary of bigger risks from AOC or similar political figures. In developed Western economies, such political messaging could gain increasing influence. Investors may think: “I don’t agree with her, but her message resonates with many ordinary people—I need to trim exposure early.”

Moderator Rob: AOC’s current odds for the 2028 Democratic nomination are ~9%. According to Arthur, this is underestimated.

Arthur Hayes:

Yes. Who else? Gavin Newsom? I think he’s burdened by California governance challenges—hard to shake free. AOC has tapped into the right message: AI capital spending drives inflation, impoverishes many U.S. communities, yet they lack financial assets to participate in the upside. She can build powerful narratives around these issues.

Moderator Andy: Illia, heard you’re now in San Francisco, having stayed there for nearly a quarter. Any observations to add? The local vibe—does it feel like the 2021 Crypto Conference, full of excitement and energy? Do you see signs of a bubble? Have you been active in local investments?

Illia Polosukhin:

Everyone is doing AI now. You’re queuing for food, and everyone’s talking about agents. It feels both like a bubble and a real transformation. Every company is pivoting toward AI—if they don’t, they’ll become obsolete.

We see this in crypto too: crypto itself is contracting, few new founders entering. So early-stage angel and seed investing is tough. AI now shows divergence: some startups raise $1B in seed rounds—making you question if your $50K investment matters. Others pursue traditional paths—building SaaS for specific use cases.

The issue: they’re competing directly against internal teams. That employee just uses an AI agent, describes needs in natural language, and the software builds it immediately. Founders build software trying to match expert-level capabilities—while experts can just describe requirements in English and let software auto-build.

So it’s hard to tell if they’ll achieve escape velocity before general tools mature. At that point, users may not need engineers to code business requirements into software.

So the space is fascinating, full of excitement—but we also see Anthropic-style companies: a team of three launching a product directly into a vertical niche. Reality: if you have a sufficiently general, horizontally scalable platform, it can be customized for anything.

When Will NEAR Become Deflationary?

Moderator Rob: Any questions you’d like to ask each other? Illia is a builder, Arthur a capital allocator. Arthur, any questions for Illia that could influence your next article’s investment thesis?

Arthur Hayes:

I’d like to ask: given the income sources you mentioned, is NEAR capable of becoming deflationary—rather than relying on in-protocol token inflation?

Illia Polosukhin:

Our goal is definitely to achieve deflation next year. We’ll continue pushing to reduce inflation.

I believe many builders in blockchain have debated: we pay too much for economic security, and costs don’t always match actual value. But from a political economy standpoint, reducing inflation isn’t easy. So we’re building governance mechanisms to make it more direct.

I’m confident we can reach a stage where, as revenue grows, lowering inflation is rational—achieving true balance. Next year, revenue will grow, inflation will fall.

Additionally, we’re exploring innovative ideas on funding security and validators—beyond traditional inflation.

Arthur’s Price Targets

Illia Polosukhin:

Arthur, do you think this market rally will remain concentrated in a few top-tier assets—or rapidly spread across the broader sector? If the spread is too fast, could it undermine the current stable landscape dominated by a few leading assets attracting capital?

Arthur Hayes:

Every vertical will have standout stars—and some underperformers. The success of top assets often justifies the existence of weaker ones. For example, in privacy and DEX spaces, you’ll find both breakout projects and weaker ones. Skilled traders can exploit this divergence for opportunities.

But as a capital allocator, I don’t want to stare at screens all day. I’ve picked my assets and will hold until macro shifts. I have price targets—like HYPE to $150. But everyone knows: if macro changes, I’ll pivot immediately.

If governments say they’ll stop funding all AI capital spending via debt—especially if those spendings fail to generate sufficient cashflows—that’s negative. If the Iran war spirals out of control, that’s negative too.

Until then, I’ll keep aggressively long. Trends are friends—until they’re not.

Compiled & Translated: DeepFlow TechFlow