When Blockchain Infrastructure Meets AI

Key Takeaways

• The blockchain industry is exploring new possibilities by developing advanced agents and tools using AI, and Zircuit aims to apply this to blockchain infrastructure to address transaction verification challenges.

• Zircuit's Sequencer Level Security (SLS) protocol enhances blockchain security by utilizing AI to identify, isolate, and validate malicious transactions.

• While improving the accuracy of AI transaction verification and increasing convenience for builders and users remain challenges for Zircuit, its integration of AI with blockchain is seen as a significant step forward in enhancing the efficiency and reliability of blockchain networks.

1. Background - Emerging Trends: AI in Blockchain

The crypto and blockchain industries have always been abuzz with trending keywords. In 2017, it was ICOs; in 2021, DeFi, NFTs, and the metaverse took center stage. This year, two standout keywords have captured the spotlight: “Meme Coins” and “AI x Blockchain.”

In the meme coin space, Dogecoin (DOGE), loved by Elon Musk and Donald Trump, has led the charge, followed by Pepe, a frog-themed meme coin, Moo Deng, a hippo-themed meme coin, and the recent sensation Peanut the Squirrel. While meme coins have been around for years and have simply gained renewed attention during this bullish phase, the fusion of AI and blockchain has emerged as a completely new theme, bringing fresh excitement to the crypto and blockchain scene.

AI x Blockchain is driving a wave of innovation, powered by the integration of advanced AI agents into blockchain systems via LLM models. Projects like Goatseus Maximus, Virtuals Protocol, and ai16z are spearheading the creation of new ecosystems. Recent efforts have included building frameworks and tooling that support the generation of diverse agents, merging practicality with meme-like appeal. A notable example is Injective's AI agent, iAgent, which simplifies blockchain tasks like payments, trading, and asset management through natural language commands.

This growing convergence of AI and blockchain, once an abstract concept, is now taking tangible shape, thanks in part to the attention garnered by AI meme coins. The impact of this fusion is increasingly visible in the form of functional products across various blockchain networks.

One notable example is Zircuit. Zircuit seeks to address challenges in operating Layer 2 blockchain networks by harnessing the power of AI. Let’s explore this case further.

2. Takeaway - Zircuit and the Sequencer Level Security Protocol

2.1 AI’s Decision-Making Surpasses Human Judgment

The defining distinction between AI and traditional mechanical or computational systems lies in their capacity for precise reasoning in specific scenarios. While logic circuits can execute decision-making algorithms, they often operate within the confines of predefined rules, limiting their versatility.

Source: AI now surpasses humans in almost all performance benchmarks | New Atlas

The advent of LLM-based AI services has demonstrated the potential of AI to surpass human judgment in accuracy across a broad spectrum of decision-making tasks. For instance, LLM-based AI services have shown moral reasoning capabilities when addressing ethical questions and outperformed humans in tasks like image classification, reading comprehension, visual reasoning, and natural language inference.

AI-based services can operate more efficiently than humans, capable of functioning continuously as long as electricity is supplied. Furthermore, AI models rely on predetermined variables and parameters for learning, reducing the likelihood of external variables such as human error significantly impacting decision-making outcomes.

Zircuit capitalizes on these strengths—AI’s higher accuracy in specific domains and reduced error likelihood—to apply these capabilities to blockchain infrastructure. By leveraging AI, Zircuit aims to address critical challenges in blockchain operations, enhancing efficiency and reliability across its network.

2.2 Limitations of Pre-Validation in Blockchain Transactions

Blockchain transactions are typically validated through pre-validation. This process identifies logical errors, such as data inconsistencies or attempts to transfer assets exceeding available balances. However, pre-validation cannot detect malicious intent behind a transaction.

For instance, in cases like the 2016 Ethereum DAO hack, if a malicious transaction is submitted to the blockchain network and contains no logical or data flaws, the network will execute it as is.

If such malicious attacks cannot be preemptively filtered by the system, the blockchain network is often forced to resort to additional measures, such as human intervention, to address the aftermath. Designing a blockchain network capable of predicting and responding to all potential malicious actions is exceedingly complex. This is why attackers continuously exploit weaknesses and flaws within blockchain systems.

Even if post-event verification reveals malicious activity, blockchain networks are left with limited options: rolling back blocks or compensating victims. These reactive measures are akin to closing the barn door after the horse has bolted, highlighting the fundamental inadequacy of traditional pre-validation in addressing malicious transactions.

2.3 Sequencer Level Security Protocol: Zircuit’s Use of AI Decision-Making

To address the aforementioned issues, Zircuit leverages AI in its transaction validation process through the Sequencer Level Security Protocol (SLS Protocol). This protocol plays a critical role in ensuring secure and efficient transaction processing within the Zircuit ecosystem.

Zircuit operates as an Ethereum Layer 2 network employing zk-rollup architecture. In rollups, the sequencer is a key module responsible for handling transactions and block creation on the L2 network. Zircuit has integrated the SLS Protocol into this sequencer to enhance its security and functionality.

The SLS Protocol enables the sequencer to pre-validate transactions using three key components: Malice Detection, Quarantine-Release, and Transaction Execution. Specifically, the Malice Detection module uses AI trained on Ethereum transaction bytecode analysis data to effectively isolate suspicious transactions. Here’s how the AI-driven SLS Protocol operates:

1. Malice Detection: An AI validation module, referred to as the Oracle, identifies transactions suspected of malicious intent and isolates them.

2. Queue for Valid Transactions: Transactions deemed non-malicious are immediately moved to the block inclusion queue.

3. Quarantine of Suspicious Transactions: Isolated transactions are held in the transaction pool until they meet specific quarantine release conditions. If these conditions are not met, the transactions remain quarantined and are eventually purged when the pool is cleaned.

4. Execution of Released Transactions: Quarantine-released transactions proceed to execution and are included in the next L2 block.

Quarantine release can occur under one of the following conditions:

• Time Criterion: After a specified period in the transaction pool, quarantined transactions are automatically released.

• Economic Criterion: The sender pays a deposit to guarantee the transaction’s legitimacy, allowing it to proceed.

• Administrative Criterion: Zircuit’s security experts manually review and approve transactions for release.

By combining AI capabilities with rule-based operations, Zircuit’s SLS Protocol effectively prevents malicious transactions from being included in the blockchain. This proactive approach significantly enhances the network's security and reliability.

2.4 Homework Assignments for Zircuit to Solve

However, there are several issues that can arise during the operation of the Zircuit network.

First, there is the possibility of AI transaction verification being flawed or excessively stringent, causing harm to legitimate transactions. To address such scenarios, Zircuit currently allows transactions to be temporarily isolated and re-verified before being included in the block. However, this approach is not a fundamental solution, as the damage incurred during the isolation period must be endured. Ultimately, this issue must be resolved by improving the accuracy of the AI model.

Another issue is that the SLS protocol requires additional effort from builders to function properly. The SLS protocol demands additional price information to detect malicious transactions and protect assets. Currently, this information is sourced from CoinGecko, meaning tokens distributed through Zircuit must be listed on CoinGecko. Additionally, if the API service providing price information malfunctions, the SLS protocol in Zircuit would also face operational difficulties.

For tokens bridged from external networks, builders must contact the Zircuit team via the official Discord channel to include them in the price system of the malicious transaction detection module. This extra step can decrease builder participation and slow ecosystem expansion. Therefore, Zircuit should adopt diversified methods for token information registration and ensure that the process is conducted fairly and transparently.

2.5 Attempts to Introduce AI at the Blockchain Infrastructure Stage

Source: Zircuit X

Efforts to integrate AI into blockchain infrastructure are still in their early stages, meaning that the technology has not been extensively explored. However, if the accuracy of implemented AI can be improved and feedback from users and builders is adequately incorporated into operations, Zircuit's attempts centered around the SLS protocol could be recognized as a valuable example. Furthermore, if the adoption of AI in blockchain infrastructure becomes widespread, resources currently consumed in this area could be redirected elsewhere, leading to the creation of a more efficient blockchain network.

The era of AI is upon us. From now on, it is expected that human judgment will increasingly struggle to match AI's decision-making and verification capabilities. From this perspective, actively employing AI in areas where it excels, as exemplified by Zircuit, could potentially mark another step forward—not just for blockchain services but also in terms of infrastructure.

3. Resource

Related Articles, News, Tweets etc. :

• https://docs.zircuit.com/