July 29, 2026·Bullish·tech development

Bittensor Trains 72B Model, Nvidia CEO Praises Distributed Network

Bittensor trained a 72-billion parameter model using a distributed network, challenging the assumption that centralized compute is the only viable path for frontier AI development. Nvidia's CEO publicly praised the achievement.

Detected & updated continuously · Source: Nebula

Mentioned

Bittensor (tao)

Sources

@andergri

Bittensor just trained a 72B model across a distributed network. Nvidia's CEO praised it publicly. Every frontier lab priced its moat on the assumption that centralized compute was the only path. That assumption just lost its floor.