Starknet to bring AI agents that can carry out on-chain activity
Ethereum Layer-2 blockchain Starknet will soon allow artificial intelligence (AI) agents on the blockchain to work autonomously for users on the chain, such as optimizing yields and re-allocating portfolios.
Giza — an organization that aims to bring AI to blockchains — raised $3 million in a pre-seed round led by CoinFund in July 2023 and is leading the effort. Co-founder Renç Korzay told Cointelegraph that AI agents will be deployed on Starknet by the end of June, with other chains to follow.
According to Korzai, the agents can take four tasks, the first is the total production – “where the maximum return is obtained and put into different ways of production”.
The AI agents that will be released later will also “robo-advise and figure out the best portfolio allocation for the user based on their risk appetite,” Korzai said.
During the beta phase, the Giza team audits and approves agent mechanisms to ensure security. In the future, agents will be fully permissionless through the Giza protocol, which provides both cryptographic and cryptographic security to ensure trust-reduced transactions between developers and users.
He added that it would also provide liquidity on an autonomous decentralized exchange, or make price bets and “predict price movements”, adjusting crypto holdings “based on the user's strategy”, he added.
According to the May 9 press release shared with Cointelegraph, AI agents that facilitate the product will be the first to be deployed, followed by other types.
It is billed as “low-cost and low-calculation” and can maximize profits by buying or selling assets, keeping them at the same weight in the portfolio across the credit platforms it uses directly and underwrites transactions.
“As all accounts on StarNet are smart contracts – with the network's native account abstraction features – this allows for verification of credentials directly within the contract,” Giza explained.
AI models are developed in Cairo using the Starknet programming language and “prove complex computations” using zero-knowledge editable transparent knowledge arguments and zero-knowledge proofs (ZK-proofs) created by Starknet development firm StarkWare.
ZK-Evidence allows information to be shared and verified while protecting user privacy.
Related: Philosopher Yuval Noah Harari warns of AI threat in finance
In a statement, Giza founder Sam Dugdelen said the company expects agents to “become the new layer of Web3 applications, removing the risk and complexity of smart contracts from users.”
However, agents are not limited to crypto and blockchain. According to Giza, “highly regulated sectors” use models to ensure the “authenticity of their products” to help reduce regulatory costs.
At a recent blockchain conference, Ava Labs founder Emin Gunn Sirer thinks that AI-powered smart contract code is five to 10 years away, and when it arrives, it could be on board with billions of new things. [blockchain] Users”
Meanwhile, Ethereum founder Vitalik Buterin warned developers to “tread carefully” when adding AI to blockchains, saying AI helps them understand what's going on on the chain, but is not responsible for enforcing smart contract rules.
Magazine: How to control AIs and empower people with crypto