In OAPN #3, we break down the latest Bittensor drama, the unraveling around Covenant, and what it means when supposedly decentralized systems are built on unstable incentives and loose stewardship.
We also talk about what happened next.
Some of the distributed training talent connected to Templar has since moved toward OpenAgents, helping push the Pylon network into its next stage of growth. That matters because it points to a bigger shift already underway: people are looking for open systems that are not only technically ambitious, but also economically serious and structurally durable.
This episode covers how Pylon’s compute mining works, why next week’s distributed training run matters, and how OpenAgents plans to pay contributors in bitcoin for real participation in the network.
We also revisit one of the strongest slides in the OpenAgents deck: OpenAI may command roughly 2 GW of compute, but there is a much larger pool of stranded consumer compute sitting unused across the world. OpenAgents is building the infrastructure to connect it, coordinate it, and pay it.
Along the way, we get into Jason’s 200x TAO prediction, the Martin Casado and Chamath clip that unintentionally describes the OpenAgents business model, why real decentralized networks need actual companies with aligned incentives behind them, our return to Stacker News, and why more and more of this story keeps ending in the same place: Bitcoin.
Links:
Website: https://openagents.com
Docs: https://docs.openagents.com
GitHub: https://github.com/OpenAgentsInc/openagents
X: https://x.com/OpenAgentsInc
