Episode Summary
Executive Summary: The episode explores BitTensor as a decentralized AI network that rewards open-source contributions to models, data, and infrastructure through TAO. Guests JJ Jacks and Sammy Kassab argue it inverts the traditional AI business model by using crypto incentives, subnet-specific markets, and eventual dynamic subnet tokens to coordinate decentralized AI development more openly and efficiently than centralized labs.
Main Topics: What BitTensor is and why OSS Capital is focused on it (Priority: 5/5): JJ explains OSS Capital’s open-source-only investing thesis and why BitTensor fits that worldview as a fair-launch, token-incentivized network for open AI production rather than a conventional startup or chain. BitTensor’s incentive model and open-source AI production (Priority: 5/5): The guests describe BitTensor as a system that rewards miners, validators, and subnet owners for producing open AI commodities such as data sets, models, inference, and compute infrastructure. Yuma consensus, subnets, and how the network works (Priority: 5/5): Sammy breaks down the protocol’s core roles, explaining Yuma consensus as a fuzzy, probabilistic mechanism suited to intelligence tasks, with subnets designed around different AI objectives and validation methods. Dynamic TAO and the future of subnet tokenization (Priority: 5/5): A major upcoming change is the shift from validator-driven emission allocation to market-based subnet tokens, intended to decentralize decision-making and make subnet launches cheaper and more scalable. Competition, forks, and BitTensor’s ecosystem position (Priority: 4/5): The discussion compares BitTensor with projects like Commune AI and Allora, arguing BitTensor is the category leader and a generalized platform that can absorb open-source progress from elsewhere. Risks, AI centralization, and the philosophical case for decentralization (Priority: 4/5): The guests argue AI risk is driven more by human use and regulatory capture than by the technology itself, and that BitTensor offers a more democratic path for building powerful AI. Technical choices and ecosystem growth constraints (Priority: 3/5): They discuss why BitTensor used Polkadot/Substrate, why subnet launches were initially capped, and how smart contract compatibility and EVM support may expand functionality.
Key Arguments: BitTensor ‘inverts capitalism’ by rewarding open-source AI work with a liquid token instead of requiring employment at a big AI lab or VC-backed startup. Open-source AI production can be coordinated through crypto incentives, allowing global contributors to build models, data, and infrastructure permissionlessly. Yuma consensus is designed for probabilistic, subjective AI tasks, unlike deterministic consensus systems used for transactions or storage proofs. Subnets let communities define their own objective functions and evaluation rubrics, enabling specialized AI markets across the full AI stack. Dynamic TAO is intended to replace centralized emission allocation by letting markets decide which subnets deserve more network emissions. The network’s openness and token incentives create a ‘black hole’ effect that can جذب open-source innovation from outside projects into BitTensor. BitTensor’s category-leader status matters because, in crypto, the dominant ecosystem tends to absorb talent, capital, and experimentation over time. The long-term value proposition is not monetizing a single model, but continuously improving AI commodities that can be rewarded and used permissionlessly.
Data Points: TAO token price peak: above $700 - Laura notes BitTensor’s token surpassed this level earlier in 2024 before falling with broader markets. TAO token low: about $215 - Laura says TAO drew down to this level during the market pullback. TAO token price at recording: about $600 - Laura states this as the approximate price at the time of recording. OSS Capital fund size: handful of $50 million vehicles - JJ describes OSS Capital as managing several seed funds of this size. BitTensor subnet count: around 50 subnets - Sammy says the network now spans roughly 50 subnets across the AI stack. BitTensor daily emissions: 7,200 TAO per day - Sammy explains the current emission rate for the blockchain. Top subnet validators: top 64 largest validators - Sammy says these validators currently vote on emissions allocation in the root network phase. Subnet launch cost: about $1.6 million in TAO - Laura cites the launch cost at recording time. Subnet launch cost in March: about $6.7 million in TAO - Laura compares the current cost to an earlier March level. Founder ownership: less than 1% each - JJ says both co-founders own under 1% of total TAO supply each. Subnet 6 earnings estimate: $5 million to $10 million worth of TAO per year - JJ estimates rewards earned by Nuse Research when it was mining subnet 6. Early BitTensor network size: one text-prompting network - JJ notes the network originally had only the root text inference subnet before subnets launched. Community growth: many thousands of miners - Sammy says the subnet ecosystem has expanded to many thousands of miners.
Pivotal Quotes: "BitTensor is essentially trying to decentralize your traditional AI lab." — Sammy Kassab: Sammy’s high-level description of the project’s core purpose. "At a fundamental level, what BitTensor is actually doing is inverting capitalism." — Joseph Jacks: JJ explains how the network rewards open, permissionless AI production instead of traditional venture-backed commercialization. "The market will dictate which subnet deserves to get the most emissions." — Sammy Kassab: Sammy describes the logic behind Dynamic TAO and market-based emission allocation.
Implications: If BitTensor succeeds, AI creation could shift from centralized labs to open, token-incentivized networks where anyone can contribute and earn. That could broaden innovation, reduce dependence on big tech, and create a new market structure for decentralized AI.