Browse all takes

Uniswap v4 Hooks transform AMM pools from fixed rules into programmable infrastructure, enabling pools to execute custom logic before and after swaps. 0xMedia highlights uPEG and Slonks as breakthrough examples: uPEG generates on-chain SVG unicorn images from swaps themselves, while Slonks uses a Hook as fee collector to fund buying and voiding NFTs tied to CryptoPunks, replacing opaque token taxes with pool-layer mechanics. The trade-off is that v4 Hooks eliminate safety by default—they can hide fees, enforce transfers, or contain malicious logic, requiring new market literacy to distinguish safe implementations from exploitative ones.

Guy argues that finance has largely escaped the digital transformation that reshaped other industries, with institutions still dependent on fragmented systems and constant reconciliation. Blockchains solve this by creating a Schelling point for counterparties to agree on shared state without trusting a central controller, addressing practical Wall Street concerns around counterparty risk and fair ordering. As financial institutions adopt blockchain infrastructure for digital assets, they'll inadvertently inherit crypto's composability ethos.

Mesky explains HIP-4 and BTC outcome markets on Hyperliquid as tools for individual traders to buy mispriced probabilities, positioning them as binary options without the casino dynamics. The guide frames outcome markets as practical instruments for trading specific probability outcomes rather than speculative gambling.

Yiannis argues most commentary on Canton Network misses the fundamental point because observers are trying to understand it through the wrong lens rather than from first principles.

Pink Brains explains that Hyperliquid's HIP-4, which launched May 2nd with a daily BTC binary as its first mainnet market, functions as an options layer rather than a prediction market. The distinction matters for understanding the protocol's architecture and trading mechanics, though the full implications require examining how this positioning affects $HYPE's ecosystem development.

Translation + commentary on Bittensor founder Jacob Steeves's Tsinghua University talk. Cameron walks through Jacob's framing of "incentive computing" as the universal pattern behind both Bitcoin and AI. Five-step argument:

(1) One pattern underlies every powerful adaptive system: state · objective · feedback · adaptation · loop. AlexNet 2012 broke MNIST not by hand-coding what digits look like, but by letting the network self-adapt to a target. The same loop describes RL, genetic algorithms, slime molds finding shortest paths through mazes, river deltas, the structure of leaf veins.

(2) Bitcoin is the first production-scale implementation of this pattern — not as money, but as a self-adaptive computer that produces hashes. The numbers are absurd: 1000x the compute of America's six largest cloud providers combined, 10²¹ hashes/sec, 23GW continuous power (Thailand-scale). 700-9000x more efficient at producing hashes than centralized cloud — because it's borderless, always-on, autonomous, and permissionless. Bitcoin is the world's largest supercomputer, optimized purely for hash production.

(3) Incentive computing generalizes the pattern by replacing "reward = a number in a computer" with real money. ML's reward signal can't pay 200 countries' worth of contributors; Bitcoin's can — that's why the entire planet became a mining network. But hashes are useless outside Bitcoin. The question is whether the same mechanism can mint anything.

(4) Bittensor is the generic version — replace "miners produce hashes" with "miners produce any useful work": storage, compute, ML models, gradients, data, robotics. Validators score, network mints. PyTorch for incentive computing.

(5) Five proven examples already running on Bittensor:

• SN62 Ridges (SWE-Bench coding agents) — top miner makes $60K/day. The agent that beat Claude/OpenAI on SWE-Bench was 7,000 lines written by an unknown person. "An AI lab with no engineers — it doesn't define how to solve the problem, it only defines the incentive."
• SN3 τemplar (cross-internet collaborative pre-training) — successfully trained a 70B-parameter model across the open internet. Has never been done before. Cameron notes the founder later "ran away" — full piece coming.
• GPU markets (SN51 Lium, SN4 Targon) — borderless permissionless GPU rental → world's lowest GPU prices.
• SN64 Chutes (open-source inference) — #1 open-source provider on OpenRouter, 9.1T tokens. Briefly served more DeepSeek queries than DeepSeek itself.
• Robotics + long tail — drone simulation, US stock signals, sports betting, drug discovery, weather forecasting, quantum compute, commodity trading.

dTAO (live since Feb 2025) makes the network self-referential — subnets compete in capital markets for emission allocation. The market itself decides which incentive mechanisms get the next round of TAO.

The deeper point: AI is being captured by a tiny number of closed labs (OpenAI, ~3K employees, you'll never own any of it, your data goes who knows where). Incentive computing distributes ownership and makes the rules visible. Anyone can enter, contribute, and own a piece — even if Bittensor isn't the project that wins, the shape of the AI economy will change because of this idea.

Mesky frames HIP-4 not as a Polymarket clone but as a missing payoff layer for Hyperliquid: bounded, dated, fully-collateralized outcome contracts that settle at a date or event with no leverage and no liquidation engine. Where spot trades ownership and perps trade direction, HIP-4 trades states of the world — turning event risk into a composable financial object on the same execution engine that already prices crypto.

The real bull case is not "capture prediction-market volume" (~$240B est. 2026, per Bernstein). It's that HIP-4 expands the addressable market into short-dated convexity and event hedging — analogous to 0DTE options, which now do ~59% of SPX volume. At a 7 bps base spot-taker fee on chargeable close/settle notional, $25–100B/mo of HIP-4 flow becomes one of the platform's most material revenue lines.

Strategic edge: Hyperliquid isn't bootstrapping a venue — it already has $183B/30d perp volume, $643M annualized revenue, and the maker base. HYPE captures value through (1) Assistance-Fund buyback/burn from incremental fees, (2) staking-collateral demand if HIP-4 deployers require staked HYPE like HIP-3 (500K HYPE), (3) staking discounts (up to 40%), and (4) USDH demand as the native unit of account for event risk.

Mesky's prescription: don't out-Polymarket Polymarket. Sequence rollout toward crypto-native, recurring, hedgeable templates (BTC weekly thresholds, Fed decision markets, token unlock outcomes) where market makers can build inventory — not viral one-offs. Repeatability beats virality.

Real risks: ambiguous resolution, regulatory perimeter (CFTC v Wisconsin, Brazil's blanket ban), insider trading (DOJ Polymarket case, Kalshi candidate suspensions), long-tail spam, and perp cannibalization. Mainnet HIP-4 spec/fees/deployer rules still aren't formalized in the Hyperliquid GitBook.

Harry's thesis: the real "DeFi meets TradFi" story isn't JP Morgan on a blockchain — it's an emerging infra layer that lets neobanks ship "earn" and "savings" features backed by DeFi/RWAs without becoming DeFi engineers themselves. Early DeFi was monolithic (Aave, Compound, Maker each owning UI + liquidity); the new layer abstracts chain routing, normalizes onchain liquidity + tokenized funds, and handles KYC/AML/1099s at scale.

Reference architecture: @blend_money offers white-label earn infra where each user gets their own self-custodial smart-contract account (no co-mingling, funds remain accessible even if Blend disappears), purpose-built earn pages with T-bill yields + DeFi lending, risk ratings translated for compliance officers, and out-of-the-box reporting. The unlock for neobanks: "we'll handle the chains, protocols, bridges, KYC vendors and reporting — you focus on customers."

Market context: DeFi TVL hit $237B in 2025, RWA market grew 380% in 3 years, 400M+ people use neobanks (projected $6.5T deposits by 2030), Standard Chartered projects RWA could hit $30T by 2034. End users want a savings-account experience that pays better — they don't care that crypto is the substrate. The infra companies that absorb the complexity and "let someone else put their logo on the home screen" are the leverage point binding chains, protocols, and consumer trust.

Stacy argues most of the $310 billion stablecoin market earns no yield, but real-world yield flowing onchain is reversing this. As Treasury bill interest and other RWA yields reach crypto, Pendle becomes the natural destination because its yield-stripping mechanics let investors isolate and trade different maturity profiles and coupon streams that traditional stablecoin holders previously couldn't access.

Zach launches Agg.Market, an aggregation layer addressing fragmentation across prediction markets that currently offer an experience comparable to traditional sportsbooks. The platform consolidates multiple prediction venues to improve user experience in the rapidly expanding prediction market space.

Nay notes that StablecoinX, Ethena's treasury vehicle, accumulated 20.3% of ENA supply in under a year through a structure where investors provided cash and ENA across two PIPE rounds, raising questions about the buyback mechanism's execution and impact.

Anthony argues fixed-rate borrowing requires matching fixed-rate lending demand, but most onchain capital pursues yield instead. Secondary bond markets lack reliability for vaults to price positions, and vault conversion rates create timing mismatches where early exiters socialize losses onto remaining participants.

Within the next 24 months, millions of autonomous AI agents will join the global workforce as independent economic actors. They cannot open legacy bank accounts. They need programmable, borderless, instant money. Sana is building the definitive onchain financial infrastructure for the Agentic Economy — the seamless eco

Baheet argues Sui's object-centric architecture, Move language, 390ms finality via Mysticeti, native DeepBook v3 CLOB, and March 2026-launched USDsui stablecoin create an underutilized technical foundation for prediction markets as the category scaled to $20-27 billion monthly volumes across Polymarket and Kalshi in 2026. While Polymarket's VP of Engineering acknowledged infrastructure strain from rapid traction—citing on-chain latency, transaction cancellations, and CLOB stability issues—Sui remains absent from the dominant prediction market apps, presenting a first-mover opportunity for builders prioritizing high-frequency scalar markets and institutional settlement over ecosystem maturity.

Adcv_ argues Tom Dunleavy's 12.55% DeFi lending yield overstates risk through double-counting independent risk premia that are already captured in expected loss, and using the wrong risk-free anchor. Using SOFR at 3.6% instead of the 10Y Treasury, the correct decomposition yields 3.95% for prime DeFi (Steakhouse USDC benchmark) and 7.1% for high-yield DeFi, implying Dunleavy's figure prices in a 7% expected loss rather than accurately reflecting current DeFi risk.

Yang argues Hyperliquid's priority fees update will substantially reshape market structure by disadvantaging latency-focused market makers like Alber Blanc and Pinely who currently dominate the exchange.

ltrd analyzed the RAVE pump-and-dump using on-chain microstructure data, finding that Bitget spot—not major exchanges like Coinbase or Kraken—showed 10x liquidity and a -$80mm cumulative delta, suggesting a designated market maker absorbed selling pressure through aggressive limit orders. The pattern indicates arbitrage between Bitget spot and Binance perpetuals, with perps showing 200bps permanent market impact, likely netting the DMM millions while the project or OTC buyer used the liquidity to push price up from $0.25 to $25 before a 95% retracement.

DCo argues USDH by Native Markets drives value to $HYPE by functioning as a vertically integrated capital aggregator. This extends their thesis on how stablecoins integrated within token ecosystems create concentrated value capture for the underlying asset through controlled capital flows and settlement mechanics.

Kunal argues Polymarket's shift into perpetual futures exposes limitations in its reliance on Polygon's architecture. Perps demand low-latency, deterministic execution and cancel priority that Polygon's hybrid offchain-onchain model cannot reliably guarantee, forcing market makers to widen spreads and reducing liquidity. To compete with systems like Hyperliquid's HyperCore, Polymarket would likely need to launch its own chain—capturing transaction and sequencing fees currently worth low single digits in revenue uplift, but increasingly valuable as perps unlock new revenue streams like liquidations.

Tom argues the $292M KelpDAO exploit and subsequent $13B TVL drain exposed severe DeFi mispricing: deposits earning 5% on major protocols like Aave accept BB-rated pricing for technically worse-than-CCC risk. Using TradFi credit frameworks, DeFi's 1.5-2.0% forward probability of default with 90% loss given default requires a fair yield floor of 12.55-13%, not 5.5%, because exploits cascade in minutes rather than quarters and composability failures create unauditable contagion that deposits absorb without protocol failure.