🔧Technical Background

To understand how TaiNet optimizes TAO staking yields beyond what is feasible for passive, individual stakers it is necessary to understand how staking and delegation works in the Bittensor ecosystem.

The Bittensor ecosystem is comprised of various subnets, which cater to different domains of machine learning. Staking rewards can be earned from subnets by delegating TAO tokens to subnet validators, thereby increasing that validator’s effective stake in the total network and increasing their total rewards.

Bittensor issues 7200 new TAO tokens each day, which are divided across every Bittensor subnet based on their performance, as determined by the Root Network - a special kind of subnet tasked with grading performance for other subnets. The largest 64 subnet validators, in terms of their stake, from amongst all the subnet validators in all the active subnets in the Bittensor network, are, by default, the validators in the Root Network. There are no network miners in the Root Network. Instead, the 32 subnets take their place. The 64 root network validators set the reward weights for the 32 subnets, mirroring the way subnet validators grade the performance of subnet miners and set their reward weights accordingly. Therefore, the total 7200 TAO created daily are distributed across the various subnets, which then distribute them further across the following ratio:

• 41% across the subnet’s miners.

• 41% across the subnet’s validators, and their delegators.

• 18% to the subnet owner.

According to this ratio, the greatest beneficiary within the subnet is the subnet owner, or any miner or validator that has ~44% or greater effective stake in the subnet - meaning they too would earn 18% or more of the total rewards distributed to the subnet.