Updated Jun 17, 2026, 1:11 p.m. Published Jun 17, 2026, 12:54 p.m.

2 min read

Pile of cash. (Emilio Takas/Unsplash)

Summary

  • A new Bittensor proposal called Root Reborn would overhaul how yield is paid to TAO stakers by turning validators into active allocators of capital across AI subnets.
  • Instead of automatically selling subnet tokens for TAO each block, validators would reinvest rewards into selected subnets, creating compounding baskets that stakers can still cash out to TAO at any time.
  • The code, currently submitted for a test network on GitHub, has already undergone an automated review that flagged two serious issues the author says are now fixed, with further cleanup planned before any mainnet release.

A new proposal for Bittensor, the decentralized AI network behind the TAO token, would change how the network pays its validators and turn them into something closer to fund managers.

Bittensor is built from dozens of subnets, each a marketplace for a different AI task, with its own token. TAO is the network's main token. Users earn yield by staking TAO to validators on the root, the layer considered the network's safest place to park capital.

The new proposal changes how yield is paid.

Right now, the system funds it by selling the rewards owed to root stakers and automatically swapping out subnet tokens for TAO. That means the network is constantly selling the very tokens its subnets are built on, which drives down their prices.

The proposal, called Root Reborn and submitted by developer 'unconst,' flips that.

Instead of selling everything, each validator would choose a set of subnets to support, much like picking holdings for a fund. The yield that would have been sold is reinvested into the chosen subnets, held as a basket that compounds over time, and staked back to the validator. Stakers still get their yield and can cash out to TAO whenever they want.

Such a mechanism stops the constant selling pressure and turns it into net buying that supports subnet prices.

Validators turn from passive yield pipes into active curators, since subnets they back attract fresh capital, while those they judge to be bad actors get starved of it.

The proposal is a code submission on Bittensor's GitHub as of Wednesday, aimed at a test network rather than the main one.

Meanwhile, an early automated review flagged two serious issues, including an upgrade step that could choke on large amounts of data and a payout path that could shortchange stakers when a subnet shuts down. The author said in a GitHub response that those issues are fixed, with more cleanup listed before any mainnet release.

Bittensor's token, TAO, has fallen 28% over the last 12 months, while bitcoin has fallen 38% over the same period. The token's staking yield currently sits around 17% if users hold TAO for a year.

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Read full story at CoinDesk