Episode Summary
Executive Summary: The September VVV call centered on Venice’s accelerating usage, tokenomics tuning, and the broader privacy backlash in AI. John argued that Venice’s token growth is the clearest signal of business health, while emission cuts, DM supply changes, and new burn mechanisms are being adjusted toward a deflationary equilibrium. The conversation also highlighted how AI labs’ data practices are driving users toward privacy-preserving alternatives like Venice.
Main Topics: Venice usage growth and token demand (Priority: 5/5): The hosts discussed Venice processing 250B daily tokens, framing token throughput as the best high-level indicator of company health and AI demand. John emphasized that usage growth is accelerating faster than earlier forecasts, with LLM activity driving most of the volume. VVV emissions reduction and deflationary design (Priority: 5/5): They reviewed the step-down in annual VVV emissions from 3M to 2.5M and then to 2M, with John saying Venice likely still needs more reductions to achieve a truly deflationary asset. The market reaction will be monitored before further adjustments. DM supply target changes and market equilibrium (Priority: 4/5): The DM target rose to 40,000, and the team is watching how staking, minting, burning, and price volatility affect the equilibrium. John said the recent price spike in VVV creates a feedback loop that can increase demand for DM as users seek to unlock staked VVV. Burn mechanisms: discretionary and programmatic (Priority: 4/5): The call explained how discretionary burns are derived from free cash flow after expenses and can be TWAPed into VVV over time, while programmatic burns are tied to product usage like subscriptions and API spend. Venice is also considering future burn expansions such as renewals. Privacy concerns in frontier AI (Priority: 5/5): A large portion of the call focused on OpenAI and Anthropic’s data retention, training, and safety practices. John argued the math-problem controversy and legal data disclosures reinforce why privacy-preserving AI is necessary and why Venice’s model is positioned differently. Minds and model orchestration (Priority: 4/5): Venice’s Minds product was presented as a bet on combining specialized models and agents rather than relying on a single giant model. The thesis is that curated, task-specific orchestration will often outperform a naked model for average users. Product and infrastructure roadmap (Priority: 3/5): John gave updates on the data center buildout, the role of Nier for high-privacy inference, and the upcoming Lumara Film Festival showcasing AI-generated film and creative tooling. These were presented as signs of Venice’s product expansion and vertical integration.
Key Arguments: Venice’s daily token count is the best single proxy for business health because it reflects real usage across app and API. AI adoption is still early; most users treat AI like “souped up Google,” while a smaller cohort is already using agents and multi-model workflows. Emission reductions are meant to move VVV toward deflationary status, but Venice will wait for market data before making additional changes. DM supply changes are not simple inflation; they are designed to create a market feedback loop where staking, unlocking, and burning respond to price incentives. Discretionary burns are effectively a free-cash-flow-based TWAP buy-and-burn mechanism layered on top of programmatic burns. The OpenAI/Codex math controversy illustrates how frontier labs can absorb user-submitted ideas into training data, reinforcing the value of privacy-first infrastructure. Anthropic’s new retention policy was framed as a safety measure, but Venice sees it as evidence that AI labs have structural incentives to retain user data. Minds reflects the belief that the future of AI value may lie in orchestration and combination of specialized models, not just a single general-purpose model. Venice’s choice not to train frontier models is a strategic advantage because it avoids the arms race that encourages data hoarding and retention. The data center buildout should improve margins, which may later feed into larger burns or other token economics adjustments.
Data Points: Venice daily tokens processed: 250 billion daily tokens - Eric Forghi’s September 16 post, discussed as a leading indicator of Venice usage growth Venice daily tokens previously: 100 billion daily tokens - Referenced as the earlier milestone that preceded the 250B figure VVV emissions: Reduced from 3 million/year to 2.5 million/year, then to 2 million/year - Token emission schedule discussed as part of moving VVV toward deflationary economics DM supply target: 40,000 - New target discussed as a mechanism to manage minting and staking dynamics Programmatic burn for Pro subscription: $2 in VVV - Burn level mentioned as the current amount burned for a Pro subscription Programmatic burn for lower tier / prior level: $1 in VVV - John noted the burn was doubled from an earlier $1 level AI usage growth forecast vs reality: 15-20% MoM expected vs 35-40% MoM actual - John contrasted internal forecast assumptions with realized growth during the year OpenAI chat logs order: 20 million chatGPT logs - A judge ordered OpenAI to turn over anonymized logs in a copyright case Anthropic data retention window: 30 days - New retention policy for covered models to support safety work Lumara Film Festival submissions: Over 700 submissions - Venice and MoonPay film festival focused on AI video creativity VBB price move on September 8: From about $18 to a peak of $30 - Referenced during the OpenAI math controversy and broader market rally VVV staking level: 100 staked VVV for lifetime program access - Mentioned as an existing utility that supports staking demand
Pivotal Quotes: "This is the best chart in crypto." — John: Describing Venice’s daily token-usage chart as the clearest measure of fundamental business health "It’s a feature, not a bug of the tokenomics." — John: Explaining how higher VVV prices can increase demand for DM as users seek to unlock staked VVV "If you don't have it, there's nothing to turn over." — John: Discussing how Venice’s privacy posture differs from AI labs that retain user data
Implications: Listeners should expect more tokenomics tuning, more product expansion, and continued privacy-driven demand for Venice. For AI users and builders, the episode argues that data retention and model orchestration will become key battlegrounds in the next phase of AI adoption.