According to Emin Gün Sirer, the founder of Avalanche blockchain, smart contract programming is a challenging task. It is difficult to capture the “intent” of the contract, coding is complex, and verification remains a challenge. Sirer believes that this is the main obstacle preventing the widespread adoption of smart contracts. However, he envisions a future where artificial intelligence (AI) agents like ChatGPT or Llama can handle smart contract coding, potentially making lawyers the primary writers of smart contracts instead of coders. Additionally, everyday people would be able to write smart contracts in their native language, making it as easy as writing a bank check.
Sirer explained this vision at the recent Cornell Blockchain Conference in New York City. He compared writing a check to the maximum complexity of transactions offered by retail banks, where you can simply transfer a specified amount of money to someone. However, in the future, he envisions adding additional instructions to such transactions, allowing for more complex smart contracts.
While Sirer believes that this future is still five to 10 years away, he believes it will be transformative and enable billions of new users to join the blockchain ecosystem. Currently, most smart contracts are written in Solidity, a computer language that is not widely known even among programmers.
Avalanche is working on building a new virtual machine that combines AI and blockchains. The virtual machine is programmed in natural language, allowing users to write programs in languages like English, German, French, Tagalog, and Chinese. However, there are still many challenges to overcome before AI agents can widely deploy these coin-operated agents.
Legal issues need to be resolved, and keywords and terms must be defined unambiguously. AI agents may not be ready to write legally binding documents yet, as they can “hallucinate” and generate inaccurate information. One potential solution is to require multiple AI agents to agree on a smart contract solution before it is accepted.
Despite these challenges, there are already some solutions available. For example, Council, an AI agent framework developed by ChainML, enables developers to provide plain English instructions that are immediately converted into executable code. Chainlink Developer Hub’s “Ask AI” can return a fully formed smart contract based on a user’s request in plain English.
However, there is still work to be done to improve the quality and complexity of AI agents writing smart contracts. They need to be trained with new content to enhance their capabilities. In the future, multiple AI agents with distinct roles could collaborate to complete the lifecycle management of a smart contract.
Sirer acknowledges that there are still many unknowns, and moving forward will require input from individuals with diverse backgrounds, including technology, ethics, and humanities. Lawyers, in particular, may play a significant role in writing smart contracts since they are skilled in using regular language to specify contracts.
Overall, this new approach to smart contracts is promising and cannot be ignored. Sirer believes that it will be built on democratic and open platforms that anyone can use on an equal footing.