Summary

Given the inherent data and computational redundancy in blockchains, deploying AI smart contracts at scale still presents many challenges and may always be inferior in performance to centralized engines. Still, some machine learning algorithms could be deployable on-chain in a cost and data-efficient manner to bring smart contracts to the next level. For the moment, the most feasible combination of blockchain and AI technologies seems to be for pooling computational resources together and distributing them across a network of algorithmic and data providers to harness these resources in a collaborative manner in order to emulate the performance and output of large, world-class AI companies.