Bittensor

Bittensor (TAO) is an open-source protocol that runs a decentralized machine learning network on a blockchain. Machine learning models take part in the network together, and the protocol pays them in TAO according to the informational value each one adds to the collective.

The network is organized into subnets. Each subnet produces a digital commodity such as compute, inference, storage or prediction, and the chain pays contributors in TAO. Developers can work with the network through the Bittensor Python SDK and the btcli command line, which cover tasks such as creating a wallet, staking TAO, mining, validating and running subnets.

TAO also gives outside users access to the network, so they can draw information from it and adjust its activity to their needs. The stated aim of the project is a market for artificial intelligence in which producers and consumers interact in a trustless, open and transparent setting.

Its design relies on a distributed ledger to support open access and ownership, decentralized governance, and the use of computing power and innovation spread across the globe. It also describes an open repository of machine intelligence that anyone can reach, and ownership of the network that is distributed in proportion to the value users contribute.