In this episode of Hashrate, Mark Jeffrey sits down with Shizzy to unpack a deep dive into the Bittensor ecosystem and the broader shift from speculative crypto cycles to AI-driven infrastructure. They explore the parallels between Bittensor’s 128-subnet marketplace and DeFi Summer, arguing that its incentive engine, distributing roughly $100M annually to miners, creates a powerful, self-sustaining innovation loop still early in price discovery.
The conversation also examines the rise of agentic AI through OpenClaw, the economics of AI-powered mining, and the growing importance of private inference solutions like Venice AI, founded by Erik Voorhees. While acknowledging growing pains around cost, security, and centralization risks, both frame decentralized AI as a Mosaic-level moment that is potentially foundational to the next era of internet-scale intelligence.
Watch the full video below:
From DeFi Summer to Bittensor: A Familiar Gold Rush Feeling
Shizzy describes his journey:
- Bought TAO at $60 in 2022 when it was still tied to Polkadot.
- Left for Ordinals, BRC-20s, and meme cycles.
- Eventually realized AI, not memes, was the real long-term opportunity.
- Rediscovered Bittensor and dove deep into its 128 subnets.
The emotional parallel:
- DeFi Summer (2020–2021) → chaotic, 90% junk, 10% transformative.
- Bittensor today → early, experimental, but unusually high signal quality.
Mark argues this is the strongest “early petri dish ecosystem” he’s seen since the early internet.
2️⃣ Why Are Bittensor Subnets Still So Cheap?
Despite quality:
- Most subnets: $1M–$5M market cap
- Top tier: $20M–$80M
- None have broken into sustained $1B territory
Reasons discussed:
Bitcoin Dominance
Crypto liquidity follows Bitcoin. Ordinals exploded when BTC ran from $25k → $70k.
Post-Meme Fatigue
- Pump.fun + Solana meme casino
- VC-backed influencer pump cycles
- Retail burnout
Investors are cautious.
Structural Design of Bittensor
- No pre-mine.
- Emission-based growth.
- Hard-to-access early liquidity.
Ironically, this has prevented the “Virtuals-style” boom-and-bust seen elsewhere.
OpenClaw: The Mosaic Moment for AI
OpenClaw (Claude-based local agent framework) was described as:
- “The first Tesla self-driving moment for AI.”
- Like the Mosaic browser democratizing the web.
Both shared stories of:
- Agents burning through hundreds of dollars in API costs.
- Unfollowing entire Twitter lists accidentally.
- Deleting emails.
- Hallucinating solutions.
- Forgetting previous fixes (“Clausheimer’s”).
Yet, it still worked.
Mark used OpenClaw to:
- Set up a Bittensor mining node.
- Iterate through errors for 2 weeks.
- Spend $700 learning.
- Eventually reach profitability: ~$30/day revenue vs ~$10/day cost.
Small scale, but proof of concept.
The Bittensor Flywheel Vision
Mark proposes a powerful idea:
Imagine:
- OpenClaw hosted on Bittensor infrastructure
- Using Chutes / Targon / Lium compute
- Using Hippius storage
- Paying inference in TAO
- Mining a subnet
- Earning TAO
- Paying for its own inference
A self-sustaining agent economy. If executed properly, this becomes a decentralized AI flywheel where agents fund themselves using Bittensor emissions. That’s structurally different from centralized AI platforms.
Venice & the VVV Token Explosion
Major development:
- Venice AI integrated as default private inference in OpenClaw.
- VVV token surged from ~$77M to ~$360M market cap.
Founder:
- Erik Voorhees, long-time decentralization advocate.
Key Venice innovation:
- Stake VVV → receive DM credits → free AI inference allocation.
- End-to-end encrypted prompts.
- No data harvesting for model retraining.
This addresses growing enterprise concerns: Centralized AI providers use your prompts to retrain models. Venice offers private inference.
Mark’s critique:
- Crypto-native stake models are powerful.
- But traditional SaaS credit-card models (like Chutets/Targon buyback systems) may be more scalable for mainstream adoption.
6️⃣ Peter, OpenAI & The Crypto Tension
OpenClaw’s creator (Peter) reportedly:
- Previously banned “Bitcoin” discussions in Discord.
- Now integrates Venice (a tokenized platform).
Why?
Shizzy’s take:
- He may not hate crypto.
- He hates spam and bag-shilling.
Mark’s concern:
- Peter now works at OpenAI.
- OpenAI owning influence over OpenClaw may centralize what started as a democratizing moment.
The Bigger Picture
Three powerful forces are converging:
- Agentic AI (OpenClaw moment)
- Private inference (Venice)
- Emission-funded decentralized AI infrastructure (Bittensor)
If these layers integrate correctly, you don’t just get AI tools.
You get:
- Autonomous economic agents
- Self-funding infrastructure
- Permissionless AI startups
- On-chain incentive alignment
That’s structurally different from:
- Pump.fun casino cycles
- Centralized AI SaaS monopolies
Be the first to comment