I'm not crazy, and the headline is not sensationalist. Let me explain everything.
Around 2008, Josh Levine (@bigjoshlevine) and others wanted to end the "conspiracy" of US exchanges, where big clients received advantageous order information, just slightly faster than ordinary users, who were unaware that their orders would contribute profits to professional teams.
Unfortunately, the trading data tool Levine created was ultimately absorbed by quantitative trading. The information was transparent, but some people's information was always a bit more transparent than others.
As a long-term FHE follower, I'm happy to see someone, especially CZ, discussing "Dark Pool", which is a potential application scenario for true encryption technologies like ZK, MPC, and FHE. However, I must say that CZ's understanding of dark pools seems not entirely consistent with blockchain-based dark pool practices.
In his narrative about dark pools, CZ focuses on hiding large order information to avoid potential targeted attacks. From the perspective of victim James Wynn, this is exactly the technology he wants. Hyperliquid's on-chain mechanism combined with information hiding functionality seems to be beckoning a more free world.
Image description: CZ comments on dark pools, image source: https://x.com/cz_binance/status/1929246168833229243
It's not entirely wrong, but CZ's idea is not exactly the same as James's experience. Let's first answer why James chose Hyperliquid over Binance for large transactions: fundamentally, he could get more and face fewer restrictions.
- No KYC, no fund review
- On-chain transaction, preventing needle insertion
Although Hyperliquid is not very decentralized, and its leverage multiples and liquidity cannot be compared with Binance, Hyperliquid's on-chain characteristics have successfully attracted whale users, which is consistent with the premise of Curve War where Curve was the only on-chain DEX used by professional traders.
However, strictly speaking, this Hyperliquid incident is not about mechanism design. In fact, the on-chain transparency is precisely why the "attacker" could see James's position and liquidation price, which is why CZ says a dark pool DEX is needed.
No, this is not the same thing. The on-chain privacy trading track has been explored to the extreme. Not to mention Arcium's FHE dark pool route, Bitcoin, Zcash, Tornado Cash, and other "true" privacy protection tools have not been mainstream market adopted. Instead, more convenient features are conquering the market.
People prefer convenience. Paying for privacy is ultimately for a very small minority.
Returning to CZ's dark pool theory, I can directly say that what he's actually talking about is "Hyperliquid after solving MEV". Coming from the traditional trading market, CZ should understand that TradFi dark pools emerged from excessive over-involvement in high-frequency quantitative trading. Large transactions can go through OTC channels, and high-frequency trading's core is competing for price advantages over ordinary users.
- Front-running or waiting for ordinary user transactions
- Predicting or attacking large transaction orders
- Being just a bit faster than high-frequency peers
The answer is clear: TradFi dark pools and blockchain MEV have similar effects. Essentially, this has nothing to do with privacy protection. However, the transparency of on-chain transactions means there's currently no real solution except "centralization". If everyone remembers, BNB Chain directly "commanded" nodes not to sandwich, barely controlling MEV.
If Hyperliquid completely eliminates MEV, seeing James's order would not manipulate the market. The only question is how we can achieve this:
- MEV elimination route
- Privacy enhancement route
Remember: Dark pools are not about "hiding transaction intent" but "hiding transaction prices", which is closer to the MEV execution process. The blockchain dark pool everyone imagines is more like a ZK/FHE-ized Hyperliquid after MEV optimization. It cannot be a CEX because centralization is the biggest MEV process.
Imagination Space Opened by Risc-V
The traditions of all deceased predecessors haunt the minds of the living like a nightmare.
I'll emphasize again that the privacy route looks similar to MEV but they are not directly related. The source of blockchain privacy technology is Bitcoin (PoW+P2P+ small amount + one-time address), while blockchain MEV and prevention technologies originate from Ethereum, where decentralized blockchain consensus mechanisms inevitably lead to inconsistencies and noise.
Even after Ethereum significantly embraces Risc-V, MEV issues will still exist. Staking nodes might increase to 2048 or even enhance super node discourse power. However, Risc-V will have new development opportunities at the software-hardware integration level, especially for FHE dark pools.
The current EVM underlying layer is not compatible with overly complex opcodes. A new VM mechanism will fundamentally change everything. This is not an alternative to SVM or Move VM, but a true innovation. In short:
- OP Code and EVM: New opcodes will support complex technologies from a lower level, improving comprehensively with one iteration.
- New VM, New Compatibility: Risc-V VM will improve from software-hardware coordination, not just software optimization.
I'll briefly mention Risc-V's significance. RAM and other architectures are chip hardware on which Windows, macOS, or HarmonyOS can be built, but they are essentially commercial products. Risc-V is an open-source chip instruction set. Ethereum can completely customize its own system, breaking free from total dependence on existing commercial hardware. Customization will bring new potential.
Returning to dark pools, we can divide them into three levels: subject anonymity, transaction anonymity, and interaction anonymity:
- Blockchain itself is an anonymous system. Unlike TradFi, anonymity is blockchain's essence. James's self-exposure or address association by data analysis platforms are probabilistic matching. Without seeing the private key, one cannot truly determine fund ownership;
- Transaction anonymity can be subdivided into price protection and order anonymity. Price protection means the price cannot be changed after determination. Order anonymity can include price protection but can also just hide the on-chain address of the transaction order without mandatorily constraining the transaction price;
- Interaction anonymity is a stage characteristic of blockchain. For example, the classic deposit/withdrawal issue: using USDT always carries the possibility of being tracked or frozen. Hackers sparing Bybit's USDT wallet illustrates this point.
From the privacy route, Risc-V has the necessary conditions for building dark pools. However, MEV issues will still exist. In current practices, ZK and TEE combination is the mainstream choice. TEE can isolate private keys, especially in multi-private key management systems. ZK can hide order details, but whether MEV can be completely eliminated depends on future developments.
In this regard, FHE might be a better technical route. Its encrypted computation characteristics can achieve subject and transaction anonymity. The only problems are high cost and slowness. With the expected Risc-V custom hardware, supporting spot DEX might be possible, but for Perp DEX, a question mark remains.
To summarize, deep integration of Risc-V and ZK technologies can provide a workable Perp DEX dark pool mechanism. FHE-specific Risc-V acceleration chips merged with Ethereum could achieve the blockchain version of the dark pool in CZ's ideal—combining anonymity, high-frequency, and large-amount transactions.
Conclusion
Whether dark pools will become reality remains to be seen, but this is clearly a capital and technology-intensive track, with only:
Historical event opportunities—CZ wants to create dark pools and knows CEX is not suitable for this;
Increasingly relaxed regulation—Tornado Cash and DeFi decriminalization mean transparent dark pools are not ridiculous. Everyone just wants safe trading, not money laundering;
Clear market demand—DEX's evolution direction is finally not about MEME. Professional traders need such tools. Curve and Hyperliquid are examples.
Satoshi Nakamoto believed in Bitcoin, so Bitcoin was born. This time, can a transparent dark pool appear in the world?
