This episode delves into the Bittensor ecosystem's potential and pitfalls, contrasting centralized AI giants with the "million drones" of specialized subnets, and exploring the critical need for user trust and accessibility in the Crypto AI space.
The Doxing Dilemma in Bittensor
- Mark Jeffrey kicks off by addressing the issue of anonymity within the Bittensor ecosystem, using Tbot (to.bot) as a prime example. Tbot offers a simplified way for users with Ethereum wallets to invest in Bittensor subnet tokens by bridging ETH to Tao within their platform, acting as a mini-centralized exchange.
- While acknowledging the convenience this provides over the more complex direct investment process via a Tao wallet and dedicated sites like TaoStats, Mark expresses significant concern over the Tbot team's lack of doxing (publicly revealing their identities).
- He highlights the inherent risk: “we don't know if we're going to wake up one morning and all the ethereum that I uploaded to their site... it might be gone.” This underscores the critical importance of due diligence for investors; anonymous teams, even with functional products, carry inherent counterparty risk, especially as the ecosystem grows and attracts potential scams.
Anonymity Among Legitimate Players
- The conversation touches upon why even known legitimate contributors within Bittensor, like the prominent subnet developer Namay, might choose pseudonymity.
- Mark speculates that security is the primary driver, referencing incidents like the kidnapping and assault of a Ledger executive. High-profile individuals known to hold significant crypto assets become targets.
- Mark suggests, “My guess is Namay just doesn't... he wants to keep his fingers.” For investors and researchers, this highlights a persistent tension in crypto between the desire for transparency and the real-world security risks faced by successful builders and holders.
Investing in Dynamic Tail (Detail) Subnet Tokens
- Mark shares his experience and evolving perspective on investing in Bittensor's subnet tokens since the "Detail" upgrade enabled easier trading.
- Initially, even experienced figures advised caution due to high initial token emissions designed to bootstrap liquidity pools on Bittensor's native Automated Market Maker (AMM) – a system facilitating decentralized token swaps.
- However, Mark argues that focusing solely on circulating market cap (not fully diluted value) reveals extremely low valuations for many subnets, especially compared to venture-backed AI startups tackling similar problems but valued in the billions.
- He states, “I look at that and I'm like we're like insane if we're not buying subnet tokens.” He draws parallels to early Ethereum, suggesting that precise market timing is less critical than recognizing the potential long-term value, even amidst volatility.
- Investors face a dilemma: enter early despite potential volatility or wait for stability, potentially missing significant upside. Mark advocates for early entry based on the vast valuation gap compared to traditional AI startups.
Evaluating Specific Subnets: Seeking Seriousness
- When asked about his preferred subnets, Mark emphasizes looking for "serious stuff." He cites Nova, a subnet focused on discovering pharmaceutical molecules in highly speculative, unfundable-by-VC search spaces, as an example of a serious project.
- He also mentions Shoots (now exceeding $30M market cap and productizing its AI) and Targon as clear contenders, alongside Proprietary Trading Network (productizing AI market prediction), Templar, Gradients (part of the Shoots network), Tensorplex Dojo (backed by Binance), and Hippus (decentralized storage, highlighted by influencer MOG).
- He contrasts these with KO (KAO), which has a large presence on X (formerly Twitter) and an existing token but faces skepticism from Bittensor OGs regarding its subnet activity and transparency.
- Mark notes KO's potential brand recognition advantage if easier Solana/Ethereum bridging emerges but holds only a "fun pile" due to uncertainties. This section provides investors with a snapshot of active subnets and the varying levels of perceived legitimacy and risk, emphasizing the need to look beyond hype towards tangible progress and backing.
Long-Term Viability: The Cathedral vs. The Bazaar
- Mark expresses strong conviction in the long-term potential of successful subnets, framing the current AI landscape as a battle between centralized "monotheistic" AI (Google, OpenAI, Grok) – the "Cathedral" – and a decentralized ecosystem of specialized AIs like Bittensor subnets – the "Bazaar."
- He references Balaji Srinivasan's shift in thinking and the impact of models like China's DeepSeek, which demonstrated that powerful AI doesn't necessarily require massive, centralized compute resources and can run efficiently, even on mobile devices.
- This leads to Mark's core thesis: “we now live in a world of a million vertically expert AIs... it's sort of like a drone swarm versus fighter jets.” He envisions a future of composable, specialized AI agents, potentially using crypto for inter-agent payments, with Bittensor positioned as key infrastructure.
- For researchers and investors, this frames Bittensor not just as a project but as a potential foundational layer for a new, decentralized AI paradigm.
Why Decentralized AI Wins
- Expanding on the Cathedral vs. Bazaar analogy, Mark argues that decentralized systems historically prevail due to adaptability and lower overhead.
- He draws parallels to the shift from mainframes to PCs ("The small drone won") and from closed online services like AOL to the open internet.
- Centralized AI giants face immense operational costs ("OpenAI is famously just incinerating cash,”) while decentralized networks are more nimble and leverage global talent.
- He compares the potential trajectory of AI to Linux, where a free, open-source core spurred innovation and commercial services built on top.
- "You can't fight the whole world,” Mark asserts, suggesting that open, decentralized models ultimately outcompete closed systems. This perspective suggests long-term strategic advantages for decentralized AI platforms investors should consider.
Evaluating Teams: Beyond Credentials to the "Kooky Edge"
- Mark downplays the importance of traditional team metrics like size or specific background history when evaluating subnets.
- Instead, he looks for seriousness and genuine commitment, often assessed through direct interaction like podcast interviews.
- He emphasizes that breakthroughs often come from unexpected places, the "kooky edges," citing Satoshi Nakamoto, Einstein, and Galileo as historical examples of outsiders driving innovation.
- "99% of it's Kookie and it's going to die... then 0.1% of it is Galileo,” he notes. Navigating this requires sifting through noise but not dismissing unconventional approaches outright, as that's where true disruption often originates.
- For investors, this means balancing skepticism with an openness to potentially transformative ideas from non-traditional sources.
The Crucial Role of User Experience and Tooling
- Mark revisits his long-standing call for better user interfaces in Bittensor.
- He praises TaoStats.io for providing the first comprehensive data dashboard but criticizes its complexity ("a space shuttle cockpit”) and Tao-centric view, making it inaccessible to mainstream users who think in dollars.
- He strongly favors newer tools like Backprop.app for its cleaner, simpler, dollar-denominated interface resembling CoinMarketCap or Uniswap, which lowers the barrier to entry.
- He stresses that simplifying the user experience is paramount for adoption. The evolution of user-friendly front ends like Backprop is a key adoption driver investors should monitor, as simplified access could unlock significant capital inflows.
Bridging Bittensor to Mainstream Liquidity
- The next crucial step, Mark argues, is enabling direct investment in subnet tokens using assets from other major chains like Solana and Ethereum.
- He mentions his upcoming interview with the team behind Sonic (formerly Fantom), a fast EVM-compatible chain, as a potential platform for such integration.
- He believes the first team to successfully build a seamless bridge ("use your Solana... connect your Solana wallet... buy subnet tokens”) will capture immense value.
- This highlights a major infrastructure bottleneck and opportunity; solving cross-chain access is vital for Bittensor's growth.
Simplicity as the Key to Mass Adoption
- Mark passionately argues that complexity is a major barrier preventing Bittensor from reaching its potential.
- He dismisses the "build it and they will come” mentality, stating it "never happens.” Using the analogy "Rome is not the marble of the Senate, it's the sand of the Coliseum," he urges the Bittensor community to focus on meeting average users where they are, rather than remaining insulated in technical complexity.
- He points to Coinbase simplifying Bitcoin access as the catalyst for its mainstream adoption and the source of Coinbase's own massive success.
- For Bittensor, achieving "product-market fit” requires descending into the "sand of the Coliseum” with simple, intuitive interfaces. The biggest opportunity might lie in building that user-friendly gateway.
Subnet Token Utility and Value Accrual
- The discussion addresses the current disconnect between subnet performance and token value, as tokens don't represent equity or direct revenue share.
- Mark notes some subnets are initiating token buybacks using generated revenue, a positive step.
- He recounts asking the Nova team about value accrual for token holders if they discover a breakthrough drug; they acknowledged the issue and are considering solutions.
- Mark suggests possibilities like direct revenue sharing (distributing Tao to token holders) or utility functions where the subnet token is required for accessing the AI service.
- He draws a powerful parallel to the PetroDollar system, where the US dollar became a "utility token for the Saudi oil app,” creating demand and value.
- This reinforces the idea that clear mechanisms for value accrual or utility are essential for sustainable token value beyond speculative interest. Investors should scrutinize how subnets plan to link token value to underlying success.
Legitimacy Through Real-World Usage
- Mark agrees that the significant real-world usage of subnets like Shoots and Targon, processing billions of AI tokens (prompts/outputs, not crypto tokens) daily, adds substantial legitimacy to Bittensor.
- The fact that users choose these decentralized options even with powerful centralized alternatives like ChatGPT available speaks volumes.
- However, he notes that this success hasn't yet translated into breakout market cap valuations recognized by the broader market.
- A major breakout success from one subnet is needed to truly capture mainstream attention.
Breakout Potential: Niche Dominance
- While skeptical that subnets like Shoots or Targon will directly "flip" giants like OpenAI in the general LLM race, Mark believes a breakout success is more likely to come from a specialized, niche application.
- He highlights ITS AI, a subnet focused on detecting AI-generated text, which benchmarks show is potentially world-leading in its specific domain.
- Potential markets like universities needing to verify student work create significant demand.
- Mark sees this as an "arms race," where both institutions (detection) and students (evasion/verification) need such tools.
- A dominant player in a critical niche could be the catalyst that validates the Bittensor model for the wider world. Researchers should look for subnets achieving best-in-class performance in specific, high-value verticals.
AI's Impact on Education and IP
- The AI detection discussion leads to broader questions about AI's role in education and intellectual property. Mark compares the current situation to the introduction of calculators in schools – initially resisted, now accepted.
- He predicts AI tools will likely become integrated into education over time.
- On IP, he discusses the controversy around AI image generation (like the Studio Ghibli style) trained on existing artists' work, acknowledging it feels like a form of "theft" or plagiarism.
- However, he complicates this by comparing it to creative borrowing throughout history (e.g., Harry Potter's structural similarities to Lord of the Rings, Star Wars blending influences).
- He strongly cautions against government intervention ("Ministry of Copyright,”) recalling early internet fears (like Sean Connery's reaction to his image online) that proved unfounded and would have stifled innovation.
- The path forward for AI and IP remains unclear but likely involves adapting norms rather than rigid, top-down control.
Navigating the Future: Humans, Robots, and AI Trends
- Asked for final advice, Mark identifies key trends. First, AI excels at augmenting humans, making them "10x” more efficient; thus, betting on skilled humans leveraging AI is a sound strategy.
- Second, AI's capability is trending towards infinity while its cost trends towards zero.
- The primary beneficiary of this, Mark argues, is robotics, particularly humanoid robots, which provide the physical embodiment ("insing”) for increasingly powerful, free AI.
- "It's making AI manifest in... steel space,” he explains. Since physical robots are finite widgets benefiting from infinitely improving, free intelligence, he sees almost limitless demand.
- His final advice for navigating the AI future: "Bet on humans and bet on humanoid robots.”
Conclusion
- The discussion highlights Bittensor's potential to foster specialized, decentralized AI, challenging centralized giants.
- For investors and researchers, the key takeaway is the critical need for user-friendly interfaces and clear token utility to unlock mainstream adoption and validate subnet valuations against traditional AI benchmarks.
