The next ByteDance is a permissionless market Media didn't just change. Its objective function did. Distribution shifted from editorial curation to engagement maximization, and the two optimize for completely different things. More than half of Americans now get news from social platforms, and the feeds ranking that news are tuned to predict what holds attention, not what is true. LLMs make this worse: they personalize at near-zero marginal cost, so the diversity of inputs any one person sees collapses toward whatever their prior already rewards. The echo chamber isn't a failure. It's the equilibrium of any system that maximizes attention. Markets are the exception, because a price is the one signal in media that costs something to produce. A market weights each opinion by conviction, and punishes error in real time. That discipline is what turns dispersed, private, tacit knowledge into public. In a media of infinite mirrors, the market is the only output that has skin in the game. But demand for that signal is thin and lumpy. Activity follows a power law: a handful of elections and headline events absorb almost all volume, while the long tail of markets sits at near-zero open interest. Two distinct failures produce this. On the demand side, the right markets never reach the right users. Distribution can't match a question to the person who actually holds an edge on it. On the supply side, even when a market exists, no market maker can afford to price it. ByteDance didn't win on better content; it won by solving the matching problem for long-tail supply. An events market feed needs the same machinery. The asset being matched is different, but the economics are identical: a vast long tail of supply that is worthless until it finds its precise audience. The supply-side failure, though, is structural. The Conditional Token Framework is clean and composable, but it pushes price discovery onto external market makers and loss-bearing LPs. Unlike perps or spot tokens, where every participant shares one deep order book and liquidity nets across the whole venue, each prediction market is a bespoke, non-fungible risk that has to be subsidized on its own. So the marginal cost of opening a market doesn't fall as the platform grows; it stays roughly constant and high. That is the inverse of software economics, and it's the real reason liquidity bootstrapping never gets cheaper no matter how many markets launch. The long tail gets excluded precisely where its information value per dollar is highest. The future of media doesn't devolve into a glorified sportsbook for lack of interest; the cost structure selects for the sportsbook. Breaking that requires a core mechanism whose engagement is invariant to notional size, where a $100 market feels as alive as a $100 million one. Without innovation at the protocol layer, every new market pays the same toll. That's the problem we're solving at 42. We're building an events market protocol that's permissionless for all creators. Pairing a suite of liquidity-agnostic mechanisms with precise distribution so that consensus can form on any topic. Long-tail topics aren't niche. They're where most tacit, local knowledge lives, and they're where prediction markets have always been theoretically strongest and practically absent. Come build the next generation of media with us. Break free from the world of infinite mirrors.

Money is going onchain. Stablecoin transaction volumes more than doubled YoY to $4.4T in Q1, and USDC now accounts for 63% of that activity [1]. We believe the infrastructure powering it is becoming one of the most important shifts in financial services in decades. That's why we're excited to announce that ARK has invested in the @arc token. For the better part of a decade, ARK's research has focused on the convergence of public blockchains, digital money, and AI. We believe @arc sits at the intersection of all three. We've watched Layer 1 networks (L1s) launch since 2015, and the space has produced remarkable innovation along the way. Each new network has contributed something, whether in consensus design, execution environments, developer experience, or economic models, and we believe there's room for several L1s with different trade-offs to have meaningful impact over the coming years. The financial system is large enough, and the use cases varied enough, that no single chain will serve every need. What a decade of observation has also taught us is that first-mover advantage in crypto is rarely decisive. @Uniswap wasn't the first decentralized exchange (DEX). @HyperliquidX wasn't the first perpetual futures DEX. @opensea wasn't the first Non-fungible token marketplace. The teams that ultimately led their categories were often the ones that arrived later, studied what came before, and built something materially better. Arc takes the best primitives developed across a decade of blockchain experimentation, including EVM (ethereum virtual machine) compatibility, sub-second finality, predictable fees, configurable privacy, and institutional validators, and combines them with something we think L1s have historically underinvested in: a strong, opinionated view on the application layer. For much of the past decade, the prevailing L1 design philosophy treated the base layer and the apps running on top as separate problems. Build a fast, neutral chain and let the market figure out the apps. That approach has produced important infrastructure, but it has also left a lot of capacity waiting for use cases to find it. @circle is taking a different path. @arc launches with USDC, EURC, USYC, CCTP (cross chain transfer protocol), Circle Mint, Agent stack and CPN (circle payment network) native from day one, with a clear thesis about who the network is for and what they'll do with it. We think that's a meaningful evolution in L1 design, and we're excited to see Circle take it on. Circle has the franchise to back it up: → $77B of USDC in circulation across 30+ chains → $21.5T of onchain USDC volume in Q1 2026 alone → CCTP processes ~60% of all cross-chain traffic → Circle Payments Network now spans 180+ countries with 136 financial institutions enrolled → 200+ ecosystem partners contributing to Arc, including Goldman Sachs, Visa, Mastercard, DTCC, BlackRock, Apollo, and Standard Chartered [2] We believe the convergence of AI and onchain finance is the single largest paradigm shift since the mobile internet. AI agents need economic infrastructure that operates at software speed. Circle's nanopayments infrastructure, which enables USDC transfers as small as $0.000001, makes machine-to-machine commerce economically viable for the first time. Arc is the rail this happens on. @jerallaire and the Circle team have spent over a decade earning the regulatory standing, institutional relationships, and operational track record that a network like Arc requires. We're proud to back this vision. The financial system is being rebuilt. @arc is positioned to be at the center of it. [1-2] Sources: Circle Earnings materials