Memory is the tightest link in the AI buildout right now: our read has DRAM at 93/100 (critical) for 276 days now When a constraint is tight, the companies that hold the scarce supply capture the rent: - SK hynix +205% vs sector (180d) - Micron +162% Early access ↓

Today, Grass announced that it is rewarding network participants in USDC instead of its native token. Grass reached profitability near the end of last year, and is now in the unusual position of being able to compensate contributors directly from the revenue the network generates. To me, this represents an important milestone. A network should eventually graduate from subsidized growth to a self-sustaining economy. Token incentives are a powerful tool for bootstrapping supply, but they should not become a permanent substitute for real demand. The end state was never "pay people with tokens forever." The end state is a business with customers, cash flow, and a network whose participants are compensated because the underlying service creates real economic value. We're still early in that journey, and there's an enormous amount of work ahead of us. But today feels like an important step, and it makes me incredibly excited for what's next.

64.4B tokens yesterday ( Priced at OpenRouter floor rates for Qwen 3.6 35B, that is ~$15,400/day. ~$5.6M/year. 100% routed to POD buybacks. Against a $15M cap, that is ~37% of the entire cap bought back annually, at the floor. And it is single-model math. The network runs a basket (Gemma 31B, 26B, cheaper and faster), so the real demand pool is wider. Buyback-to-emissions: 7.6x. Net supply shrinks as it scales. For scale: Venice runs ~100B tokens/day at a $616M cap. Dolphin is already at 64% of that throughput for under 3% of the valuation, and it builds the uncensored models that power the Venice ecosystem. Same category, real partnership. $156M FDV is not an overhang when the protocol buys back faster than it emits. And the API is not even live yet. The buybacks have not started. This is the floor. Pre revenue, pre OpenRouter listing, pre agent demand. base:0xed664536023d8e4b1640c394777d34abaff1df8f

Have seen a bunch of people pushing back on the Anthropic news being a positive catalyst for Venice user growth and $VVV with variations of “well it’s not like Venice has a model that can compete with Mythos.” That’s entirely missing the forest for the trees. Sure that’s the first order effect but the second order effect is the impact on user data and privacy. If one of the top AI labs is willing to switch off access to users at a moments notice, what will prevent them from sharing all of your personal data or gating models/features based on user profiles (location, nationality, background, etc). It’s not about what other models does Venice offer that can compete with Mythos today but rather when you use Venice, usage is private/anonymized/uncensored. The other implication of the Anthropic news is the clear need for decentralized training and open source models that can one day compete near the levels of the models produced by the frontier labs. That’s a long term problem to solve but there are also a lot of interesting protocols working on this, some with liquid tokens but many still pre token launch.

not sure if btc continues to provide diminishing returns but it is certainly possible (i choose to believe it wont for now) but alts (on average) will continue to get more difficult and dangerous to trade imo, irreversible trend. (1) massive competition amongst increasingly sophisticated buyers (who simultaneously believe less) (2) primarily traded on perps/with leverage. (3) huge dilution in coins with mkt valuing stuff high by default without justification (4) launch FDVs always capturing 100% of optimism for the asset without respecting price/valuarion (5) too much pre-market price discovery for (4) to be safe the average 2017 buyer buys spot and hodls weeks/months cos they believe, add on the way up, asset was trading at low val early so works out. average 2025 buyer is buying on perps without checking the valuation and sells whenever their PNL goes red or force sells in liquidation. however, it will remain the best place for returns for smart ppl. skill expression and asset selection is much more important. patience much more highly rewarded over being "early" on liquid markets last few years. and then outlier assets will continue to exist, maybe 1 every couple of years, and when they turn up you can turn brain off and 2017 it. imo anyways (hopefully)