Westfield, NJ, January 19, 2026, Chainwire
Crunch’s decentralized ML intelligence layer opens up Bittensor mining to a community of over 11,000 ML engineers and 1,200 PhDs.
Crunch plans to allow coordinators to mine Bittensor subnets and manage the technical infrastructure while the community focuses on model development. Crunch’s approach retains Bittensor’s decentralized spirit while abstracting the complexity of the blockchain for contributors who are excluded by some of the coordination overhead. Rather than hiring existing crypto-native miners, the platform onboards entirely new participants from enterprise and academic ML backgrounds, expanding the diversity of intelligence flowing into the subnet without requiring contributors to be familiar with blockchain.
Bittensor is a decentralized open-source blockchain network that is building a global market for artificial intelligence (AI) by encouraging users to share their computing power, data, and AI models. Mining on Bittensor means providing computational resources and machine learning models to different subnets of the network, specialized AI services that compete to solve specific problems.
Bitensor’s subnet architecture is a highly compatible place to deploy collective intelligence. Therefore, mining on the revenue-generating Bittensor subnet is a natural fit for the Crunch community. This makes Bittensor’s subnets another class of “customers” for Crunch collective intelligence.
“The majority of the Crunch community is data scientists and ML engineers from corporate and academic backgrounds who want to contribute to decentralized AI without becoming blockchain experts. Our role is to bring together compatible opportunities and abstract some of the complexity of decentralization so they can focus purely on what they do best: building better models,” said Jean Herelle, founder of CrunchDAO.
Bittensor subnet metamodeling layer
Crunch’s metamodeling approach leverages the principles of collective intelligence, where ensemble models built from submissions from diverse contributors consistently outperform individual predictions.
By aggregating models from hundreds of independent ML engineers, Crunch creates a robust intelligence stream that leverages the wisdom of the crowd. There, the combined forecasts from various contributors often exceed the forecasts of individual experts. This collective intelligence method translates directly into Bittensor subnet mining, where crunch coordinators aggregate community submissions into ensemble models to provide more reliable and diverse intelligence to the network. A subnet creator or coordinator can purchase mining slots and launch a modeling market that effectively functions as a metalayer, or “superminer.”
This move addresses a critical talent bottleneck in decentralized AI. ML scientists at traditional academic institutions and institutions have deep expertise in model development, but the complexities of Web3, such as managing mining slots, staking mechanisms, and blockchain infrastructure, have largely excluded this talent pool from participating. This exclusion comes at a cost. Decentralized AI offers ML researchers things they can’t do in centralized labs: economic ownership, open experimentation, and global awareness tied directly to real-world model performance.
By abstracting blockchain complexity through Crunch’s coordinator infrastructure, these scientists can focus purely on what they do best: building great models. This will create an additional revenue stream for ML practitioners while bringing much-needed specialized intelligence to the decentralized ecosystem, addressing the critical talent shortage in decentralized AI.
By bridging academic AI expertise and decentralized infrastructure, Crunch enables a new generation of contributors to access other intelligence networks without compromising decentralization.
Coordinator participation
Subnet operators interested in launching coordinator nodes can contact coordinators@crunchdao.com.
More information about the Crunch Protocol Testnet, as well as updates and community discussions, are available through Discord and X.
About CrunchDAO
Crunch is a distributed machine learning intelligence layer that connects businesses and institutions with more than 11,000 data scientists and engineers to solve predictive challenges across industries, from quantitative finance to genomics. Through the Open Modeling Marketplace, organizations have access to thousands of competing models that are evaluated purely on performance, and contributors earn rewards based on their results. With over 35,000 models deployed and proven clients including ADIA Lab and the Broad Institute of MIT and Harvard University, Crunch makes elite AI expertise fluid, transparent, and accessible.
contact
CMO
Andrew Zapo
Crunch Lab Co., Ltd.
andrew.zapo@crunchdao.com
