The Bittensor Bull Thesis with James Ross and Mog

For today's episode, we dive deep into the the intersection of AI and crypto, and discuss how this sector has produced more hype than results (so far).

While both fields promise decentralization, combining them effectively has proven challenging. Most projects either centralize AI decision-making or create token systems without clear utility.

Bittensor is taking a different approach to this problem by creating a network where AI models compete for rewards through market incentives. This system lets token holders decide which models receive rewards based on measurable performance, rather than promises or marketing.

Their recent update shifted voting power from validators to token holders, aiming to align incentives more directly with utility. One project testing this approach is Mode, with their new subnet Synth, an AI model competing to predict market volatility.

The Synth subnet is live, with data scientists building rival prediction models. Models face off in open competition with market-driven feedback, creating natural selection—strong performers earn more rewards, while weaker ones get filtered out.

Let’s dive into what this means, how Synth is applying it, and what’s next for onchain AI.

The Rollup