Is Hyperliquid's original sin still "centralization"? The wall-hitting dilemma of on-chain DEX.

Hyperliquid voluntarily removed the trading pair after the JELLY attack, which is an extremely "centralized" ultimate crisis handling behavior in the eyes of some communities, and has become an original sin of "saving or not saving" on DEX. (Synopsis: Hyperliquid's "price manipulation" actively closed JELLY's short orders were inflamed, Arthur Hayes: Bet that $HYPE will fall back to square one) (Background supplement: Hyperliquid Fright Night" giant whale played "malicious bursting" official once lost tens of millions of magnesium, capital flight dangerously broke roots) Last night, the crypto community was clamoring "HYPER needs to be revalued." The reason is that Hyperliquid, which claims to be a high-performance on-chain derivatives exchange, once again reported that its HLP liquidity vault (Vault) was attacked, resulting in about millions of dollars being severely cut by market manipulators shorting left hand and right hand spot. This incident has once again pushed DeFi to a fundamentally acute question: when the infrastructure of a "decentralized" exchange is actually highly controlled by a single team, where is the line between it and "centralized"? Is the (current) best security mechanism actually people? Hyperliquid may be, or may be, the epitome of the dilemma faced by many on-chain DEXs (decentralized exchanges) when challenging the supremacy of CEX (centralized exchanges). Review: Well-Designed Market Manipulation The market operation on Hyper last night was not about exploiting traditional smart contract vulnerabilities. The attackers appear to have targeted Hyperliquid's HLP vault mechanism. The vault is similar to GMX's GLP, allowing users to deposit a portfolio of assets (such as stablecoins, ETH, BTC, etc.) to acquire HLP tokens as counterparties to platform traders, sharing transaction fees and profits and losses. The crux of the matter is how the HLP price is calculated. Attackers artificially distort the "Mark Price" of these assets by performing extreme operations on certain relatively illiquid trading pairs on the Hyperliquid platform (for example, investing a large amount of money to pull up or smash the market in a short period of time). Since the HLP's net worth calculation relies on the marker price of the assets it holds, this price distortion causes the valuation of the HLP to be pulled up sharply in an instant. Subsequently, the attackers used the "inflated" HLP in their hands as collateral to lend other assets (such as stablecoins) on the Hyperliquid platform that far exceeded their actual value, and eventually transferred these assets away, leaving behind inflated HLPs and actual asset losses, which were ultimately borne by other liquidity providers in the HLP vault. The losses caused by the JellyJelly incident are estimated at around $4 million, and if there is no official compensation, these losses are ostensibly hanging on the heads of depositors. The appearance of "decentralization" and the "centralized" core Hyperliquid is a high-performance DEX built on its own Layer 1 blockchain "Hyperliquid L1", which aims to solve the problem of slow and high cost of the Ethereum mainnet DEX. In theory, this is a technical path to greater efficiency and user experience, and it can also solve some of the regulatory problems that CEX will face. However, the big players who pursue market manipulation have played hi from CEX, how can they let go of this new paradise? In order to achieve the high throughput and low latency it claims, Hyperliquid L1's current network validator is only run by the official core team. This means that although transaction settlement takes place on the blockchain, the ordering, verification, and even state changes of the entire chain are actually in the hands of a single entity, which seems to be very "centralized". This "centralized, decentralized" model brings several concerns: If there is a problem with the Hyperliquid team's servers or infrastructure, the entire trading platform could come to a standstill. It also gives teams the ability to selectively process transactions, and even roll back or intervene in extreme cases (although there is currently no evidence that they will). When the storm hits, users must trust the Hyperliquid team not to do evil and not abuse control of private chains and protocols. This is essentially the same as CEX's need to trust exchange operators. Even CZ often shouts that transparency brings trust, not to mention that Hyper has just stepped on the position of DEX, and it will take longer to stabilize its footing, take Binance to benchmark, the larger the market, the easier it is to be scolded. WHILE THE DIRECT CAUSE OF THIS JELLY MARKET MANIPULATION INCIDENT IS THE VULNERABILITY OF ORACLES (OR MARKER PRICE CALCULATORS), THE COMMUNITY IS CURRENTLY POKING AT THE CENTRALIZED VALIDATOR STRUCTURE BEHIND IT, WHICH RAISES ANOTHER QUESTION: IF THE NETWORK IS REALLY CONTROLLED BY A SINGLE TEAM, WHY CAN'T ANOMALIES BE DETECTED FASTER, INTERVENED TO PREVENT, AND EVEN INTERVENE IN FAVOR OF USERS IF NECESSARY? THE EXISTENCE OF THIS CENTRALIZED CONTROL HAS PUT HYPER IN THE FACE OF A CRISIS IN AN EMBARRASSING SITUATION WHERE IT CAN NOT COMPLETELY REMOVE RESPONSIBILITY (BECAUSE IT HAS CONTROL TO DIRECTLY PULL OUT JELLY FROM THE SHELF), AND MAY NOT BE ABLE TO STOP LOSSES IN TIME BECAUSE THE RESPONSE IS NOT "CENTRALIZED" ENOUGH (LOOK AT A HACKED CEX REACTION AND PUBLIC RELATIONS IS REALLY FIRST-CLASS). DEX slammed into the wall, why is it so hard to shake CEX? Hyperliquid's dilemma is not unique, it reflects the challenges that are prevalent in the current DEX competition with CEX: User experience (UX) and ease of use: CEX provides integrated services, from fiat deposits and withdrawals, spot trading, derivatives to wealth management products, usually with a friendly interface and a low barrier to entry. DEX requires users to manage wallets, private keys, understand gas fees, cross-chain bridging and other concepts, which is not friendly to newbies. Liquidity and trading depth: The top CEX brings together a large number of users and market makers around the world, with excellent liquidity and trading depth, and low slippage. DEX'S LIQUIDITY IS RELATIVELY SCATTERED ON DIFFERENT PROTOCOLS AND CHAINS, ESPECIALLY FOR NON-MAINSTREAM COINS, THE DEPTH IS OFTEN INSUFFICIENT, AND THE SLIPPAGE OF LARGE TRANSACTIONS IS HIGH, AND JELLY HAS BEEN USED FIERCELY THIS TIME. Performance and cost: While Layer 2 and dedicated application chains such as Hyperliquid L1 attempt to address performance issues, there is still a gap in efficiency compared to the CEX centralized matching engine. At the same time, on-chain interactions inevitably incur gas fees (which exist even on L2). Security risks: The main risks of CEX lie in the security of the platform itself (hacking, insider evil) and hosting risks. In addition to the possibility of phishing on the front end, DEX also faces multiple on-chain native risks such as smart contract vulnerabilities, price oracle manipulation, flash loan attacks, and economic model design flaws. As the Hyperliquid incident exposed, even if the contract itself is not vulnerable, attacks around its AMM mechanism can cause huge losses. Hyperliquid and the "application chain DEX" model it represents try to find a balance between performance and decentralization, or just don't say anything, just connect the traditional CEX computer room to the chain, like when the POS working mechanism was just popular, many people still laughed at the "computer room chain". Once you encounter something like this JELL...