{
  "id": 12188102,
  "title": "I run Bittensor miners. Here's what paying for measured output actually does.",
  "url": "https://urgent.news/2026/10/05/i-run-bittensor-miners-heres-what-paying-for-measured-output-actually",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-05T16:31:24.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/dato_b/i-run-bittensor-miners-heres-what-paying-for-measured-output-actually-does-1iki"
  },
  "original_language": "en",
  "account": "This reporter operates Bittensor miners, not as an investment but as an infrastructure operator. They compete daily with other operators to secure work assigned by the subnet's scoring code. This experience has altered their perspective on AI quality control. Bittensor functions as a market where machine intelligence is organized, with each subnet defining a task and paying the best-performing operators in TAO tokens. The dynamic-TAO upgrade introduced individual tokens and economies for each subnet, with TAO holders determining the flow of value. Here are five key insights from their experience:\n\n1. Payment based on measurement can shape behavior. Subnets are only as good as their scoring rule; operators tend to optimize their work to match what the scoring code measures, potentially compromising the task's original purpose. The most successful subnets continuously refine the alignment between measured output and the value it generates.\n\n2. Inactive or non-functional subnets are routinely removed. In June 2026, Bittensor's leadership began halting emissions for inactive or non-functional subnets, which became a permanent weekly policy by July. Dozens of subnets have been cut since, with operators treating this policy as an expected part of the ecosystem rather than a dramatic event.\n\n3. Vesting schedules influence the quality of work submitted. In the earned-media subnet, operators commit to publishing content before receiving payment in daily installments over a month. If the published article is deleted or converted into an advertisement, the remaining installments are forfeited. This back-loaded payment structure encourages operators to consider the long-term viability of their work, rather than focusing solely on short-term gains.\n\n4. Decentralized entry does not guarantee equitable outcomes. Despite Bittensor's decentralized nature, some operators capture a disproportionate share of mining rewards on certain subnets. While open doors do not ensure a fair distribution of winnings, they do prevent individuals from locking others out entirely. This distinction is often overlooked in discussions about decentralized AI systems.\n\n5. The public nature of the scoring board is invaluable. The reporter holds some TAO and subnet tokens and finds great value in the public visibility of who produces, who stops, and who gets paid. This transparency allows for easy verification of claims and provides a level of accountability that is not typically found in closed AI labs. The question at the heart of this experiment is whether incentive engineering can reliably produce quality at scale, and the answer appears to be a resounding yes, as long as the scoring rule allows for it.",
  "summary": "I operate miner neurons on several Bittensor subnets. Not as an investment, as an operator. I run infrastructure that competes daily against other operators for work that a subnet's scoring code decides how to pay for. A year of doing this has changed how I think about AI quality control. Not because of anything anyone promised, but because of what the payment schedule does to behavior, mine…",
  "key_points": [
    "Payment based on measurement can shape behavior in Bittensor subnets.",
    "Inactive subnets are routinely removed to maintain ecosystem health.",
    "Vesting schedules encourage operators to consider long-term viability."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}