Bittensor breakout fuels AI token rally as distributed training gains credibility
A rally in the AI token market is driven by Bittensor's success in distributed training, demonstrating a viable alternative to centralized AI models.Summary
The GMCI AI Index has seen a 48% increase since February, though it remains significantly below its all-time high. This rally is largely attributed to Bittensor (TAO), which comprises nearly 25% of the index and has nearly doubled in price in March. The surge in TAO’s value is linked to the demonstration of credible distributed training capabilities with the Covenant-72B model, a 72-billion-parameter language model trained across 70+ nodes, achieving competitive results on the MMLU benchmark. This achievement challenges the notion that distributed training is too slow or fragmented. Further bolstering the ecosystem, the Bittensor subnet ecosystem is maturing, with Targon securing a deal to power Dippy AI. Currently, over 10.7 million TAO tokens are issued, with over 68% staked, and the price stands at approximately $317, with a market cap exceeding $3 billion. The development has also attracted attention from industry figures like Jensen Huang and Chamath Palihapitiya.
(Source：The Block)