Exchange Ranking
Top Exchanges
24h Volume Ranking
Popularity Ranking
Exchange BTC Balance
Proof of Reserves
Decentralized Exchanges
Funding Rate
Funding Heatmap
Liquidation Data
Max Pain
Long/Short Ratio
Whale L/S Ratio
Binance/Okex/Huobi L/S
Bitfinex Margin L/S
ETF Tracker
Solana ETF
XRP ETF
Hong Kong ETF
Bitcoin Treasuries
Crypto Reversal
Ethereum Reserves
HyperLiquid Wallet Analysis
Hyperliquid Whale Watch
Large Transactions
On-chain Movement
Bitcoin ROI
Stablecoin Market Cap
Options Analysis
News
Articles
Economic Calendar
Features
Wallet
Contract Calculator
Security
Collections
Watchlist
Following
ETH Vitalik Buterin believes AI will help cyber defenders
Ethereum co-founder Vitalik Buterin refuted the idea that "stronger AI will make cyber security undefensible." He believes that artificial intelligence can also help developers build software with stronger mathematical security guarantees. He pointed out that both attackers and defenders will receive more powerful AI tools, making software design and verification increasingly important.
Butrin said that when developers use formal verification more widely, cybersecurity may favor the defender. Formal verification uses mathematical methods to prove that software follows established rules. He argued that while artificial intelligence helps attackers find vulnerabilities, advanced AI can also make the generation of these mathematical proofs cheaper and easier.
There is an increasingly common view that AI hacking technology means that network security is hopeless. But I disagree. I believe that once people clarify their thinking and standardize their operations, cybersecurity is inherently beneficial to the defender. Anyone who continues to own cryptocurrency (including me, who accounts for approximately 90% of my net worth) is implicitly assuming that developers can create digital systems that are resilient to increasingly powerful attackers. [Source: Twitter/X]
他将这种可能性与AI系统解决数学问题的案例进行了类比，例如费马大定理和纳维-斯托克斯方程。他的核心观点是，人工智能可能会将软件安全性声明视为需要证明的数学命题。
形式化验证仍面临定义难题
布特林指出，最艰巨的任务可能在于定义什么是“安全的软件”。他以Signal应用为例：攻击者可以针对消息传递、密钥发现、服务器、操作系统、库、编译器、数据库或硬件发起攻击，而加密仅能保护消息内容本身。
这给形式化验证带来了巨大的挑战。证明只能确认开发者所定义的规则。如果这些规则遗漏了某种威胁，证明就无法覆盖该风险。布特林表示，因此安全定义可能变得极其冗长，需要在AI进行数学证明之前进行细致且严谨的工作。
以太坊安全工作迈向更强的证明机制
最近的以太坊开发工作也显示出对基于证明的安全性的持续投入。布特林支持EIP-8288提案，该提案旨在通过递归STARK聚合来组合密码学证明，从而降低量子安全交易的成本。该提案目前仍处于草案阶段，目标是为未来的以太坊升级做准备。
以太坊开发人员也在辩论未来的交易系统应如何平衡规模、隐私和抗审查性。近期关于账户抽象标准的工作显示，以太坊Layer 1和Base在网络会议结束后采取了不同的设计方案。这场辩论反映了布特林所描述的更广泛的挑战：开发人员必须首先明确安全目标，工具才能对其进行验证。
