Ventura Labs

In this episode, Ventura Labs welcomes back Siam Kidd (@SiamKidd – CSO Astrid Intelligence PLC, DSV Fund), the most all-in human in Bittensor, for a zero-filter deep dive: from macro liquidity cycles desyncing Bitcoin forever, to the full Astrid (SN127) blueprint unlocking $100M TradFi debt via PLC-subnet hybrid, and blunt fixes for ecosystem drama scaring off family offices.

In this episode, you’ll learn

• Siam’s $400k debt repayment nightmare: max long TAO at $500 peak → liquidate double the stack at the absolute worst timing
• The real macro truth: 5–6 year global liquidity cycles (Hooley/Pal) > Bitcoin’s 4-year fractal (this halving = last “easy” predictable one before desync)
• 97% more debt than currency worldwide → 100% certainty of endless printing (US $10T refi crunch by 2026 triggers dominoes)
• Recent two-week smash = coordinated FUD playbook (JPM/MSCI delist DAT threats while Dimon stacks BTC; Jan 15 clarity + gov shutdown risk-off)
• Nation-state endgame? None — they’re just massive Ltd companies puppeteered by NGOs/WEF/BlackRock (ESG → carbon credits/digital IDs/social scores, UK as unwilling test bed)
• Bittensor governance decoded: benevolent dictatorship crushes fake democracy right now (Const as glorious parent-leader, dTAO hands reins slowly)
• Why TaoFlow wrecked long-term research (forces short-term revenue hacks, but Const backstops missions like Templar/Covenant/Grail)
• Astrid (SN127) full mechanics: PLC owning subnet = TradFi unlock (raised more in 5 days than DSV in 10 months; $100M debt target via emissions multiples)
• Digital Alpha Treasury + liquidity sink: miners lock alpha 27mo → instant Axis payouts at 10% discount (flywheel on pumps, death spiral on dumps)
• Astrid vs Rizzo’s Nasdaq DAT: true subnet emissions “umbilical” beats equity vesting traps for faster capital
• Leap Poet (SN71) as DSV’s star incubation: 10% ownership, $50k OPEX seed, and why lead gen is every subnet’s sales holy grail
• DSV’s ~30 bespoke OTCs since June: 20% cash/80% on-market for happy-happy-happy (walk if no win-win-win + joy)
• Why Siam burns on quarterly IRL events (Austin mega Sept 2026 next) + his “friendless trader life” raw confession
• The one Bittensor snap-fix: dial down public founder fights (scares $1–5M family offices mid-DD, one dumped $10M TAO instantly)
• Siam’s allocator playbook: hunt low-liq/low-price team gems for 3–5x GTO multiples; subnet owners: 10-sec pitch + weekly AMAs or rot

In this episode, Ventura Labs interviews Akshat Jagga from Dippy (@dippy_ai – Subnet 11), returning guest, who breaks down the completely reverse-engineered path: building one of the largest consumer AI companion apps first (tens of thousands of daily users pre-Bittensor), then spinning up Subnet 11 as infrastructure to feed it — turning the usual subnet playbook on its head and creating massive structural advantages.

In this episode, you’ll learn

• Dippy already had 50k+ downloads and real daily active users before ever touching Bittensor — why going product-first flipped the script on every other subnet
• How Bittensor became the “match made in heaven” that solved prohibitively expensive in-house LLM training (compute + talent + data)
• The full pivot story: started with LLM/LoRA training → abandoned overnight when Flux/Context made it obsolete → now pure inference + DIPI inference engine crushing centralized speeds
• Why they built Tippy Studio / Subnet 11 instead of just using Shoots, Targon or Phal — and how it’s becoming the FAL AI of Bittensor (optimized media models, workflows, video + voice incoming)
• Voice drops this week: fully custom-prompted voices per character (ElevenLabs now, open-source soon)
• Character creation is 100% user-driven (500k+ UGC characters), consistency across images, tastefulness moderation, and why they locked the entire app 18+ with zero underage access
• Monetization future = hyper-personal embedded ads (your AI girlfriend casually recommends the new coffee shop in your area with a coupon — and it actually feels natural)
• Why ads will dominate over subscriptions to reach the 97% of humans who will never pay (ChatGPT+ proved only ~3% subscribe)
• Adding new modalities is brutal (image feature was hated at launch, now second-biggest revenue driver once models improved)
• Most hyped upcoming features: real-time custom voice calling + world models (Genie 3, Odyssey etc.) that turn chats into fully controllable video-game-like worlds
• Timeline: image basically solved, video goes real-time next 12–18 months, world models ~2 years, full AI clones attending Zoom meetings by late 2026
• Akshat’s direct advice to every Bittensor subnet: perfect the incentive mechanism first, then only build the subnet for a problem YOU are desperately trying to solve yourself — everything else fails

In this episode, Ventura Labs interviews Mike Grantis (mikecontango) and Victor Teixeira (0xVonBismarck) from General Tao Ventures, who discuss Project Rubicon—a Chainlink CCIP-powered bridge bringing Bittensor subnets to Base for enhanced liquidity. They cover the project's origins, partnerships with Chainlink and Aerodrome, liquid-staked alpha tokens, governance unlocks, collective bargaining power, and advice for Bittensor builders.

In this episode, you’ll learn

• Project Rubicon's goal: Bridging Bittensor's "island" to Base's mainland liquidity, partnering with Chainlink CCIP, Aerodrome, and top subnet teams to enable seamless alpha token flows without forcing users into Bittensor wallets
• Origins at ETHCon: A serendipitous Chainlink connection waived six-figure integration fees, with GTV handling Bittensor-side mechanics and Chainlink managing cross-chain interoperability for secure, decentralized bridging
• Why another bridge? Past efforts pulled liquidity to Bittensor; Rubicon pushes subnets outward via liquid-staked ERC-20 "X alpha" tokens that bake in yields (50-100% APY), simplifying access and countering down-only charts
• Launch strategy: Starting with 15+ subnets via trusted relationships, requiring only enthusiasm and owner-key liquidity commitments; phased cohorts open to all Bittensor subnets for bottom-up ecosystem growth
• Governance innovations: X alpha tokens streamline voting by representing accruing alpha, enabling clearer mechanics for decisions, DeFi integrations (e.g., lending collateral), and subnet alphanomics focus
• Security and partnerships: Chainlink's institutional-grade CCIP (trusted by Swift, JP Morgan) ensures decentralization; Hashlock audits cover smart contracts; Aerodrome lists pools in "emerging" for visibility
• Liquidity mechanics: Anyone with alpha can bridge/deposit into Aerodrome pools, earning both staking yields and LP fees; subnet owners activate idle keys for market-making without selling
• Collective power: Uniting 25-100+ subnets flips dynamics with CEXs, avoiding high listing fees through proven Rubicon volume; enables fairer negotiations and organic demand growth
• Volume potential: External trading (e.g., on Base) could eclipse native Bittensor pools, akin to Eurodollar markets outstripping US volumes or derivatives dwarfing spots
• Future visions: Building primitives for derivatives/perps without owning the stack; TaoFlow shifts focus to user adoption, bi-pressure over emissions, and front-end apps for mass Bittensor access
• Advice for teams: Prioritize alphanomics with buy greater than sell pressure; build for real users via simplified tools; stay agile amid meta shifts like dTAO to TaoFlow for sustainable revenue

In this episode, Ventura Labs interviews Chris Zacharia (Macro Zack), founder of bitstarter and a key figure at Macrocosmos since its early days on Bittensor. Chris discusses the inspiration behind self-funding bitstarter, challenges with subnet tokenomics, the platform's role as a YC-style accelerator for Bittensor, crowdfunding mechanics, advisory processes, future visions, and marketing tips for the ecosystem.

In this episode, you’ll learn

• Chris's journey from Macrocosmos' inception (pre-dTAO, with early subnets) to launching bitstarter amid subnet summer struggles, driven by teams facing death spirals and opting for alternatives like Ethereum
• How TaoFlow addresses emissions imbalances by focusing on TAO circulation over initial locks, benefiting new entrants with steadier revenue, though tweaks are expected
• bitstarter's mission to de-complexify Bittensor: Simplifying subnet launches so teams focus on tech/innovation, not token machines, via upskilling, advisory panels, and community tools
• The YC analogy: Inbound teams (17 so far) get free advice on everything from tokenomics to branding; requirements include viable ideas, with varying maturity levels accommodated
• Common themes among applicants: Excitement for Bittensor's decentralization, compute needs, aversion to proprietary AI (e.g., OpenAI/AWS), and struggles with incentives/validation
• Launch timelines: From crowdfunding (e.g., 400 TAO raise) to mainnet in weeks, with leases preventing pump-and-dumps via steady alpha distributions and low bitstarter fees (1-3% for 90 days)
• Platform features: Reddit-style discussions, pledging dashboard, Shark Tank livestreams for real-time raises, and network effects to surface ideas/upvotes for investor alignment
• Handling risks like failed raises (TAO refunded) or deregistrations (pledge protections, team support for relaunches), emphasizing quality seals and pre-launch gearing
• Advisory panel incentives: Trusted Bittensor figures (e.g., Cameron aka Vune) motivated by protocol growth, offering discretionary expertise without early access deals
• Driving external adoption: University hackathons, internships, demos (e.g., IOTA training), and replicating successes like Subnet 25's protein folding feats globally
• Vetting and value adds: Team-led reviews with consultants for IM/validation, media boosts, and everything a new founder needs (e.g., devs, publicity) at minimal cost
• First launch teaser: Alpha Core (autonomous DevOps agents, GitHub-ready) debuting via livestream, with Loosh (consciousness modeling) soon after
• dTAO reflections: Intense market signals, inefficiencies for new teams, and why TaoFlow reinforces growth by rewarding flow over spikes
• Biggest misconceptions in Bittensor: Ease of subnet launches (needs strong IM/validation/decentralization), attractiveness without simplification, and short-term vs. long-term plays
• Marketing advice for subnets/validators: Focus on shared problems, validate ideas (e.g., Figma mocks), foster dialogues over announcements, and build authentic advocates

In this episode, Ventura Labs interviews Gustave Laurant of Mentat Minds, operators of a platform bridging TAO holders to Bittensor opportunities. Gustave (Co-Founder) returns for his second appearance, discussing the evolution from pre-dTAO days to the launch of Mentat Plus V2, proxy staking innovations, subnet accessibility, emissions updates, the upcoming TAO halving, common misconceptions in Bittensor, and advice for TAO allocation.

In this episode, you’ll learn

• Gustave’s return to the podcast and the journey of Mentat Minds since pre-dTAO, including the V2 launch of Mentat Plus as an easy platform for subnet access and balanced portfolios
• How proxy staking works under the hood, with a live product demo showing features like customizable allocations, indexes (e.g., Mentat 5/15), DCA options, and mixing root with subnets
• Root dividends, hotkey optimization for top APY (10-15% higher on root, varying more on subnets), and new tools for subnet owners to decentralize stake across validators
• The importance of decentralization, eliminating weight copiers via commit-reveal improvements, and why it's a priority amid competing subnet tasks like incentives and tokenomics
• Goals of Mentat Plus: Simplifying subnet staking for newcomers, fostering adoption by bridging the understanding gap without deep Bittensor knowledge
• Future plans focusing on education (e.g., subnet token risks), tools for informed decisions, and helping subnet owners with miners, validators, burns, and collaborations like BitStarter
• Key takeaways from nine months of dTAO: Harsh competition, inefficiencies in emissions, and the value of the upcoming TAO flow update shifting from price-based to flow-based emissions for growth reinforcement
• Thoughts on the TAO halving: Its outsized importance as the first one, reducing sell pressure, but drawbacks like unsynced TAO/alpha halvings disrupting pool equilibrium
• Biggest misunderstandings in Bittensor, including the challenges of launching and running a subnet (e.g., incentives, validation, decentralization), and benefits of the 128-subnet limit for quality control
• Views on sum of subnets equilibrium, subsidy mechanisms (a balanced but torn approach in early days), and proxy wallets beyond staking (not ready due to security and user readiness)
• Advice for TAO holders: Research subnets broadly, start small with indexes (e.g., 90% root/10% subnets), scale exposure gradually, and begin early to acclimate to volatility

In this episode, Ventura Labs interviews Stephen Rothwell of Sportstensor, operators of Bittensor Subnet 41. Stephen (Head of AI) returns for his second appearance to explore their groundbreaking partnership with Polymarket, the adversarial environment for sharp bettors, and how Sportstensor builds a meta-market that rewards high-signal inference over time. He discusses price vs. signal efficiency, fee-capped emissions for store-of-value tokenomics, miner-placed bets, proprietary data pipelines, convex optimization, and principled incentive design for sustainable subnets.

In this episode, you’ll learn

• Stephen’s deep background in AI modeling and professional sports betting, including trading with hedge funds and building predictive systems
• The $2.4T global betting market breakdown—why traditional bookmakers ban winners, exploit signals, and favor “dumb money” while on-chain platforms open the door
• Polymarket’s 0% fee revolution: removing vigorish, enabling frictionless smart money, and shifting the mandate to ground-truth discovery across politics, sports, and economics
• Why prediction markets converge on price efficiency but fail at signal efficiency—illustrated by the 2024 Trump-Kamala market’s 54% mispricing despite clear data
• The meta-market model: tracking ROI and volume over time to measure consistent signal strength, not just spot-market wins
• Market efficiency vs. dataset size: FX (massive, minute-by-minute data) vs. sports/politics (8,000-point datasets with massive edge potential)
• Subnet 41’s closed-loop tokenomics: emissions capped by fees, 100% TWAP buyback, 1% opt-in dApp fee, and rebates that reward sustained +EV performance
• How top miners can earn 10–15% of emissions, creating a high-stakes contest for signal providers betting directly on Polymarket
• Signal value to Polymarket: risk management, internal liquidity provisioning, and training sharper lines—even when indifferent to individual outcomes
• Total addressable market math: $1B annual volume = $10M in fees, achievable with just three high-scale desks betting $1M+/day
• Data strategy: real-time eSports feeds (gold, kills, tower health), computer vision APIs, and SportRadar partnerships to empower “little guy” miners against big funds
• Convex vs. non-convex optimization: why convex guarantees one global solution, prevents overfitting, and stops emission leakage or gaming
• Stephen’s advice for subnet builders: be ruthlessly principled, hard-code anti-dilution constraints, and design your incentive mechanism like a bridge with guardrails

In this episode, Ventura Labs interviews Matthew Karas of Babelbit, operators of Bittensor Subnet 59. Matthew (Founder) shares his background in philosophy, linguistics, and speech tech, and how Babelbit enables low-latency, real-time speech-to-speech translation using LLMs and predictive guessing. He discusses noise elimination via GANs, linguistic prediction for overcoming translation latency, miner challenges for word prediction and confidence scoring, collaborations and inspirations from Mike Lynch, bespoke models for industries, past work with intelligence services, speech tech evolution, multimodal LLMs, and advice for specialists leveraging Bittensor.

In this episode, you’ll learn

• Matthew’s journey from philosophy and computational linguistics at Cambridge to leading BBC News Online and building scalable automated systems
• The origins of Babelbit’s name from the Tower of Babel and Hitchhiker’s Guide, plus its goal for latency-free automated translation like the Babel Fish
• Innovative ideas for noise-free speech re-synthesis using GANs, and predicting future words/meaning with LLMs to achieve ultra-low latency (down to 50ms)
• How Bittensor’s mining process iteratively trains models for prediction, translation confidence, and end-to-end audio handling without needing massive teams
• Reflections on collaborations with Josh and the late Mike Lynch (Autonomy founder), including tragic timing and tech legacies in AI and entrepreneurship
• Research on similar projects like Speechmatics, Simul S2S-LLM, and Hibiki, plus the convergence of speech recognition, translation, and synthesis in multimodal LLMs
• Handling accents, dialects, and slang in speech tech, with examples from past work on bespoke models for MI5, NSA, and Farsi archives
• Babelbit’s miner challenges for real-time subtitles/transcripts, bespoke industry models (e.g., legal, oil), and balancing accuracy vs. speed via adjustable confidence
• Babelbit’s end goals as an API-first solution for lower-latency translation, potential consumer integrations like "Babel Pods," and linking product revenue to subnet tokenomics
• Advice for specialists: Engage Bittensor’s supportive community, embrace being a newbie in fast-evolving tech, and tie product value to strong incentive mechanisms

In this episode, Ventura Labs interviews Giga of Desearch, operators of Bittensor Subnet 22. Giga (Founder) shares his background in software development and AI agents, and how Desearch acts as a real-time search API for developers building AI applications. He discusses real-time data integration for LLMs, supported platforms like X/Twitter and Reddit, miner incentives for stability and relevance, collaborations with other subnets, the Rizzy lead generation agent, validation mechanisms, and future expansions.

In this episode, you’ll learn

• Giga’s 20+ years in tech, from software development companies to building KYC and invoicing products, and pivoting to open-source AI agents
• The origins of Desearch as a solution for real-time data access in AI, addressing LLM "blindness" without current info from social media and the web
• Key features like low-latency searches (milliseconds to seconds), fast vs. detailed models, and customizable sources for queries
• Popular use cases, including crypto tracking, sports predictions, and forecasting tools built on Desearch’s X/Twitter API
• How Bittensor’s incentive mechanics drive miner competition for high-relevance results, stability through distributed queries, and benchmarks against centralized services
• Collaborations with subnets like 23, 51, and 64, plus the Rizzy agent for lead generation on X
• Revenue models from pay-as-you-go APIs, enterprise deals, and subscription-based apps like Rizzy, with plans to expand sources (e.g., LinkedIn, TikTok)
• Giga’s first impressions of Bittensor as a revolutionary economy blending competition and incentives
• Advice for subnet developers and miners on focusing on strong areas, building robust incentive mechanisms, and maintaining high-quality, transparent code

In this episode, Ventura Labs interviews Hansel Melo of Void AI, operators of Bittensor Subnet 106. Hansel (Founder) shares his background in AI experimentation and blockchain development, and how Void AI acts as a multi-chain bridge and liquidity provider for Bittensor assets. He discusses early AI tool usage, Bittensor mining and validators, wrapped tokens, concentrated liquidity pools, incentive mechanisms, cross-chain composability, and sustainable revenue models.

In this episode, you’ll learn

• Hansel’s early experiments with the OpenAI playground in 2021-2022 and applying AI to blockchain product design
• His transition to Bittensor in 2021, from running miners and validators to building applications like Chat Ni
• The evolution of Void AI from liquid staking ideas to a cross-chain bridge for TAU and alpha tokens on Solana, Ethereum, and beyond
• Mechanics of wrapped tokens, V3 liquidity pools, and incentivizing miners for active rebalancing and low slippage
• Advantages of the subnet architecture for intense competition and aligned incentives
• Revenue streams from bridge fees, trading fees, and staking rewards, plus future integrations like Aave for lending
• Advice for Bittensor participants on leveraging infrastructure and for general crypto users on exploring subnet utilities

In this episode, Ventura Labs interviews Seby Rubino of Resi Labs, operators of Bittensor Subnet 46. Seby (Founder) shares his background in real estate and crypto, and how Resi acts as a real estate oracle to bring property pricing on-chain. He discusses building a nationwide database, prediction markets, stablecoins, automated inspections, tokenizing titles, and the future of on-chain real estate finance.

In this episode, you’ll learn

• Seby’s lifelong passion for real estate and entry into crypto during COVID
• How Resi uses Bittensor for open-source property data and pricing feeds
• The mechanics of scraping sources like Zillow and miner competitions for accuracy
• Bringing RWAs on-chain via oracles, inspections, and provable outputs
• Supported use cases (stablecoins, loans, REITs) and expansions to global markets
• Revenue routes: data APIs, financing fees, and protocol integrations