An unprecedented milestone is currently capturing the attention of @nvidia CEO Jensen Huang and @chamath in their recent conversations. They are discussing the most extensive decentralized pre-training event for a large language model ever recorded in history. Working through SN3, @tplr_ai successfully completed the training of Covenant-72B by collaborating with 70+ contributors across open internet infrastructure. This exciting achievement clearly signals that distributed, open-weight model training on Bittensor is truly just beginning its journey.