• Track VVV for continued growth in Venice usage, but wait for evidence that token growth is translating into revenue and sustained burns before treating it as a high-conviction buy.
• Watch the scheduled late-October reduction in VVV annual issuance from 2.5 million to 2 million, while treating further cuts and a shift to deflation as uncertain.
• Reassess VVV after Venice reports on its first data center, expected around late October or November; its effect on margins may take time to evaluate.
• Treat DM as speculative: monitor staking, unstaking, and mint-and-burn activity over the next few months, since the new 40,000 supply target has no established long-term effects.

Venice (VVV)

• Venice reported processing 250 billion LLM tokens per day on September 16, up from 100 billion on June 25. The team described usage growth as faster than expected: roughly 35–40% month over month this year, compared with earlier forecasts of 15–20%.
• A Venice strategist estimated that 1 trillion daily tokens could be reached around January or February, while saying that a December milestone was possible but less likely. This was an estimate, not a formal target.
• The team’s stated goal is for VVV to become deflationary. Annual issuance was scheduled to fall from 2.5 million to 2 million VVV in late October. The strategist suggested 2 million may still be higher than needed, but said no further reduction had been decided.
• VVV burns come from subscriptions, API-credit purchases, and discretionary team-directed purchases. Discretionary burns are funded from a portion of free cash flow and occur over time; the team said they may eventually become less important relative to programmatic burns.
• The transcript cited VVV rising from $18 to $30 on September 8, alongside heightened attention to AI privacy issues. This is historical price action, not a price target.
• Staking VVV can provide access to Venice features, including lifetime Pro access at 100 staked VVV, and can be used to mint DM.

Takeaways

• Venice’s rising token usage is a potential indicator of increasing product demand, but token volume alone does not establish profitability or guarantee that VVV’s value will rise.
• Monitor actual revenue, programmatic and discretionary burns, issuance changes, and the effect of the company’s planned data-center buildout on margins. The team said the first data center was expected around late October or November, with financial effects likely to take time to assess.
• Treat future emission reductions and burn levels as uncertain: the team said it would evaluate market responses before making additional decisions. The transcript also notes that rapid price moves can affect staking and tokenomics.

DM (Venice’s DM token)

• Venice raised its DM supply target to 40,000. The team said it was monitoring how the change affected staking, minting, burning, and the amount of VVV locked.
• The strategist said increased locking had accelerated after earlier target changes, but had recently eased, in part because some VVV holders wanted to unstake. The team described the data as noisy and said it could take a few months to better understand the change’s longer-term effects.
• Under the described mechanics, holders who want to unlock VVV may need to burn DM. This can create demand to buy DM when holders seek to unstake; minting and selling VVV can create the opposite pressure.
• The team said DM’s price had risen despite increases in its supply target, but emphasized that the system was still developing and had limited data.

Takeaways

• Follow DM supply changes alongside VVV staking and unstaking activity; the transcript does not establish a stable long-term equilibrium.
• The mint-and-burn mechanism may link DM demand to VVV price movements, but it can also make the relationship more volatile. The discussion offered no DM price target.

OpenAI and Anthropic (private AI companies; potential IPOs)

• The speakers said IPOs from major AI labs could come at the end of the year or early the following year, but gave no confirmed timing, valuation, or investment recommendation.
• They argued that demand for AI inference is growing as models become more capable and users run more complex tasks.
• The discussion highlighted privacy concerns: OpenAI could not rule out that user-submitted research had entered its training data, and the speakers cited a court order requiring OpenAI to turn over 20 million anonymized ChatGPT logs.
• Anthropic was described as changing its data-retention policy for certain enterprise arrangements from zero retention to 30 days for covered models, citing safety work. The speakers questioned whether this could reduce user trust.
• The speakers characterized the AI industry as an intense competitive race in which labs have incentives to gather data and improve models.

Takeaways

• The transcript presents AI as a fast-growing sector, but it does not provide a way to value either company or specify an investable security. Any IPO access and terms would need to be assessed when officially announced.
• Privacy, data handling, legal demands, and competition are material issues raised in the discussion that could affect user trust and business practices.

NEAR (NEAR)

• Venice said it uses NEAR as a provider for some of its trusted execution environment (TEE) and end-to-end encrypted inference options.
• The team described these services as a relatively small part of Venice’s overall inference, but important for users seeking cryptographic privacy assurances. Such options can be more limited or expensive, depending on the model.

Takeaways

• NEAR’s connection to Venice is presented as an infrastructure and privacy-provider relationship, not as a major source of Venice’s overall usage.
• The transcript gives no NEAR price target or specific investment recommendation. Assess the broader NEAR ecosystem separately from this partnership.

AI inference, privacy, and decentralized access

• The speakers described rising AI usage as a potential investment theme and framed Venice as a way for crypto users to gain exposure to AI inference economics.
• They contrasted centralized AI labs that build large general-purpose models with services that combine or route between specialized models. Venice’s Minds product, described as being in beta, aims to package model combinations for specific tasks.
• Venice said it does not currently train its own foundation models, instead integrating models developed elsewhere. The team argued that this avoids the cost and pressure of competing directly in model training.
• The speakers also discussed AI-generated film and video as a creator opportunity, citing a Venice–MoonPay film festival and more than 700 submissions. They said AI could lower production costs, but did not quantify the economics.

Takeaways

• Track whether privacy-focused AI services and model-orchestration products convert growing AI interest into sustained users, revenue, and token burns.
• The transcript makes a case for potential growth in AI applications and creative tools, but provides no specific valuation, return estimate, or timeline for those themes.
• Privacy concerns and data-retention practices are central to Venice’s positioning; continued growth would depend in part on whether users value those features enough to choose the service.