Starknet, an Ethereum layer-2 blockchain, is gearing up to introduce AI agents capable of autonomously executing on-chain tasks like optimizing yield and adjusting portfolios for users.
Giza, a company focused on integrating AI into blockchain systems, secured a $3 million pre-seed funding round led by CoinFund in July 2023 and is leading this initiative. Renç Korzay, one of Giza's co-founders, informed that AI agents would be deployed on Starknet by the end of June, with plans to extend their capabilities to other chains later on.
These AI agents are designed to perform four key activities initially, including yield aggregation to identify the highest return opportunities and allocate funds accordingly.
Subsequent releases will see them acting as robo-advisors, determining optimal portfolio allocations based on users' risk preferences.
During the beta phase, Giza will oversee and approve agent strategies to ensure security. Eventually, through the Giza Protocol, agents will transition to a permissionless model, offering cryptographic and cryptoeconomic security for decentralized interaction between developers and users.
Furthermore, these agents can provide liquidity on decentralized exchanges, execute leveraged trades, and predict price movements to adjust holdings based on user-defined strategies.
T
he first wave of AI agents will focus on yield optimization, with additional functionalities to be introduced later, according to a press release issued on May 9.
Utilizing the Starknet programming language Cairo, the AI models employ Zero-Knowledge Scalable Transparent Argument of Knowledge and zero-knowledge proofs (ZK-proofs) developed by StarkWare to verify complex computations while preserving user privacy.
Giza co-founder Cem Dagdelen envisions these agents becoming the new application layer for Web3, simplifying smart contract interactions for users.
While the potential applications extend beyond crypto and blockchain, Giza suggests that AI-powered models could assist highly regulated sectors in reducing oversight costs by validating outputs.
Although some industry experts anticipate AI-backed smart contract coding to become mainstream in the next five to 10 years, Ethereum co-founder Vitalik Buterin urges caution, emphasizing that while AI can aid in understanding on-chain activities, it shouldn't be entrusted with enforcing smart contract rules.
May 2024, Cryptoniteuae