Revenue Search: Inside Bittensor

Victor Teixeira (General TAO Ventures) traces his path from Contango Digital’s $10M blockchain–AI fund to becoming a full-stack Bittensor operator — incubating or advising multiple subnets (e.g., 23, Red Team, TPN), running the Round Table validator, and mining across the stack. He spotlights Subnet 35 (“Cartha”), an FX perp DEX co-built with Taoshi (Subnet 8): miners are either LPs or trader-miners; trading generates fees of which 50% go to LP miners and 40% go to vote-escrowed alpha holders as USDC dividends (weekly), creating aligned “alphanomics” that reduce sell pressure and reward real usage.

The broader discussion centres on revenue-first sustainability in the post-DTAO world. Buyback-and-burn isn’t dismissed, but Victor argues subnets should prioritise tangible earnings, robust VE models, and even prefunding/VC capital to bridge idea→revenue—plus new tools like lending against owner keys. With the TAO halving likely to thin liquidity growth per block, they expect more volatility (especially for smaller pools) and a premium on subnets that can fund OPEX, miner/validator sell-side, and sustain top-leaderboard emissions via real customers. For guest wish-lists to bring fresh eyeballs into BitTensor, Victor suggests Barry Silbert and Ejaz (Bankless)—and generally “everyone,” because mainstream adoption arrives when users benefit from Bittensor under the hood without even noticing.

Gopher (ex-Masa) / Subnet 42 provides cryptographically verified, AI-ready data via a network of ~256 miners running trusted execution environments to scrape and normalise web/X/Reddit/TikTok plus large financial price feeds. Devs use a UI + API (credit-based billing) and a built-in vector DB to aggregate/search topics and power apps. Traction: ~77k users across products (≈54k on AI Insights; ≈20k on a new trading tool), ~$1M ARR, ~1k paying on the trading app launched ~2 weeks ago. That app ingests multi-timeframe price data, generates trade setups, and can execute on Hyperliquid; team claims ~65% win rate with ~1:4 risk-reward on internal accounts and will publish on-chain trading wallets.

Strategy & token alignment: Gopher is migrating Masa to a Cosmos L1 (“Gopher”) focused on data aggregation/apps (Q1 launch target). Subnet 42 remains the data engine; enterprise and app customers pay fiat credits, and usage-based revenue from Gopher’s stack flows to SN42 for the data it supplies. The plan is for Alpha holders on SN42 to govern revenue use (buybacks, treasury, growth, etc.), keeping value with the miners who create it. Near-term focus: expand financial feeds (~100k assets), improve low-latency delivery (~26 ms), add social/news signals into trading, and explore a fund/vaults that trade the signals—while continuing to court market makers, prop funds, and other subnets that need dependable, verifiable data.

Vidaio (Subnet 85) does AI video upscaling (SD → HD/4K) and compression (dramatically smaller files with similar perceived quality). Their consumer web app is live in beta—demo showed ~95% size reduction—with paid tiers coming (think ~$0.05/min, ~75% margins). Beyond creators, the big targets are streamers/broadcasters, legacy libraries, security/medical, and autonomous fleets—anyone drowning in storage/CDN costs. Near-term roadmap adds a streaming pipeline (auto encoding ladders + Hippius storage), plus R&D on colorization and selective video generation/inpainting.

For enterprises needing NDAs and tighter control, Vidaio introduced an Enterprise Track: vetted “Elite miners” execute jobs off-subnet; clients pay fiat split roughly ~75% to miners / ~25% to Vidaio. Miners must post an AlphaBond—locking ~50% of their fiat payout equivalent in ALPHA until client acceptance—creating ALPHA demand/lockups while miners get fiat to cover infra (less sell pressure). Vidaio’s cut first covers OPEX, with surplus flexed between buybacks, product, and growth. Benchmark goals: surpass Topaz Labs (~$8.3M/yr consumer) and Visionular (~$10.5M/yr enterprise) while keeping the subnet as the innovation engine.

AWS went down, so we kicked it old-school on Zoom with Seby (RESI, Subnet 46). Quick refresher: RESI is building a real-estate oracle—unlocking U.S. property data and on-chain intelligence for lenders, DeFi/fractional RE, and proptech. Seby walked through fresh updates: a new white paper and alphanomics, DSV added to their OTC stack, V2 moves from APIs to scraping (cheaper for validators), V3 splits the subnet into data, inference, and storage (with Hippius) so prompts like “price this home from its inspection” fetch comps, analyze reports, and persist results. A public dashboard is rolling out, and go-to-market leans on IDX-style white-label portals for brokerages to drive viral distribution.

On revenue, RESI’s already selling Predict CASA data packs via a paid funnel (Meta ads → two-call close) and tightening the machine with hires and higher spend; the Oracle appraisals target fractionalized/DeFi real estate at $500 setup + $100/mo. Alphanomics is pure-alpha (no new token): stake for pricing tiers, LP with alpha for deeper discounts, and a planned sidechain where gas/wrapped-alpha and LP incentives amplify buy pressure as builders launch on top. TL;DR—Resi’s shipping product, lining up customers, and scaling a sales engine to convert leads into buybacks while turning a locked market into composable, on-chain real-estate intelligence.

Mark and Siam sit down with Jill Kenney, CEO of sundae_bar (Subnet 121)—a publicly listed (AIM: ESBAR) AI agent marketplace that lets developers list, host, and monetise agents while businesses buy ready-to-use or customised agents for workflows (HR, sales, ops, etc.). The twist:tsundae_bar uses its BitTensor subnet as a supply engine—briefs (from users/miners) are voted on, miners build against them, validators/humans quality-check, and top agents get boosted in sundae_bar's search. Payments are live, early agents/templates (e.g., from LETA and NAT) can be deployed now, and the platform is builder-agnostic (they aim to integrate more builders like AgentKit once stable). For enterprises needing ring-fenced deployments, bespoke off-market builds are on the roadmap.

The business model is classic marketplace: a take rate on sales, ads/featured placement, and hosting/analytics fees—with a portion earmarked for alpha buybacks and an employee alpha treasury, reinforcing the subnet’s emissions flywheel. Because sundae_bar is a PLC, it can raise capital to accelerate growth while the subnet channels top dev talent into the store. Big picture: make agents dead-simple to find, test, and deploy so a solo founder or a KPMG-scale team can stitch together a digital workforce—today via templates, soon via end-to-end customizable agents that “just work.”

Mark and Siam sit down with Austin, founder of Aurelius (SN37)—an AI-alignment subnet built on Bittensor. In plain English: training gives models knowledge; alignment adds wisdom. Aurelius tackles the “alignment faking” problem by decentralising how alignment data is created and judged. Miners red-team models to generate high-resolution synthetic alignment data; validators score it against a living “constitution” (beyond simple Helpful-Honest-Harmless), aiming to pierce the model’s latent space and reliably shape behaviour. The goal is to package enterprise-grade, fine-tuning datasets (think safer, less hallucinatory chatbots and agents), publish results, and prove uplift—then sell into enterprises and researchers while exploring a token-gated data marketplace and governance over the evolving constitutions.

They cover why this matters (AGI timelines shrinking, opaque lab pipelines), what’s hard (verifying real inference, building a market), and how BitTensor gives an edge (cheap, diversified data generation vs centralised labs). Near-term: ship a proof-of-concept dataset, harden LLM-as-judge, expand integrations (Shoots/Targon), and stand up public benchmarks (Hugging Face, peer-reviewed studies). Longer-term: Aurelius as a decentralised “alignment watchdog” layer that continuously stress-tests frontier models and nudges them toward human values—so the future’s smartest systems aren’t just powerful, but prudent.

BitQuant (Subnet 15 by Open Gradient) wants to be your on-chain “pocket financial advisor”—chat in plain English, get analysis, and execute swaps across supported chains—all powered by a BitTensor subnet of competing miners.

Guests: Matthew (ex-Two Sigma) and Advait (ex-UCL founder; PhD track) from Open Gradient.

Their bigger vision: an L1 (Open Gradient) for verifiable, privacy-preserving AI agents (TEE + zkML) running on-chain. Their BitTensor product line is BitQuant (SN15): miners compete to produce higher-quality quantitative analysis and trade guidance that the frontend at bitquant.io turns into simple, chat-driven actions (discover, analyze, swap, manage risk).

Mark & Siam clear the rumour mill: they have helped broker Subnet 71 for a new team (funding + intros + “stabilisers”), official comms coming from the team soon.

Guest Sam (Subnet 10) then lays out SN10’s mission: kill the onboarding friction. Today, a typical newcomer wrestles with wallets, TAO, bridges, and hours of steps. SN10’s cross-chain swap fixes that.

In a live Revenue Search special, Shak (Ridges) explains how they’ll shift incentives from benchmarks to real user impact: the product itself will decide who earns emissions. Ridges V1 ships as a Cursor/VS Code extension on Oct 30, 2025, priced around $12/mo (with an opt-in data tier near $8).

Under the hood, validators still run SWEBench/Polyglot, but an additional step silently swaps in challenger agents for a slice of users; miners get paid only if those users accept more suggestions, need fewer fixes, and stay engaged.

Recent mixed-set scores dropped from ~88% to ~17–18% when Polyglot was added, then rebounded to ~41% by Oct 6—evidence, Shaq says, that iteration speed is their edge. A full platform rewrite lands this week (stability, parallel evals, dual-sandbox on device, limited internet excluding benchmark content) and USD payouts are returning to attract company-scale competitors.

Goal: grow users fast, reach revenue > emissions (targeting by January) to both disincentivize gaming and potentially fund buybacks—while remaining far cheaper than rivals.

In this episode, Revenue Search digs into Subnet 41’s Sportstensor with Leo and Stephen (Neuromancer).

Hosted by Mark and Siam, this episode unpacks how their new mechanism rewards only winning, conviction-backed flow routed to prediction markets. They lay out a Polymarket partnership where Sportstensor builds a layer on top and charges a 1% fee on traded volume, using those fees to buy back alpha and, if needed, burn it. The incentive design is anti dilutive - miners never receive more in alpha than the fees generated by their qualified volume. The team cites real results from last year’s models, plus how they previously pushed roughly half a million dollars of volume to Polymarket. Expect plain talk on guard rails, why tiny or reckless bets do not count, and why opening mining to skilled traders beyond Bittensor matters. It is a candid strategy session on aligning incentives with truth seeking markets.

• NBA season result cited: 14% ROI, with MLB averaging about 5 - 6% ROI
• Partnership detail: route trades to Polymarket, charge 1% on volume, fees fund alpha buybacks and burns
• Mechanism claim: anti dilutive payouts capped by fees, with proof over promises and skin in the game only