1-shotted with opus 5.5 the assembly line of how @USDai_Official product is evolving over time: with GLX and computeiq added in as we become more TVL-invariant, we will be measured on velocity and throughput

The Copernican Model Over the last few months, I’ve been studying the crypto models that seem to be discussed on a "fundamentals" basis: HYPE/Aave/Venice/etc, following the users/ revenue/relationship between a useful product and its token - all to create a sort of "value accural ontology." This is a case study from that research. The reason was more selfish: I wanted to understand which models work when your business has to interact with the "real world," namely for how we want to take @usdai_official for the most alignemnt As a founder, I’ve had doors literally shut on my face when we started because we’re “crypto” (literally left in the waiting room like a tv show). I still believe tokenization is one of the most important technologies we have today. But the institutions we need to work with often want a equity counterparty, a board, and a contract. You just wont get around this for the next 10 years (no a cayman foundation doesnt work) That experience has made me increasingly interested in the role of an equity surrogate: an independent operating business that connects a network to the world outside it.. but frankly there werent many clean examples out there without value destruction. The case I keep returning to is @tradexyz and Hyperliquid.. and im not really sure they even know how symbiotic of a relationship they created. They technically SHOULD be the same team.. but they are divided across the river due to regulatory issues, while being symbiotic Consider the S&P 500 agreement, they licensed the benchmark to tradexyz for a perpetual market: the institutional relationship sits with the company, but the product runs on the network. tradexyz can negotiate licenses, develop products, operate markets, and build distribution. Hyperliquid provides shared execution and settlement infrastructure. And if theres a company that ever wants "exposure" to the Hyperliquid business, but they just cant buy a "token" (and believe me, it's super super hard to get a traditional enterprise to buy a token, it doesnt fucking matter if you have an ETF, you literally have to convince a boomer board to do it and get the infrastructure set up to do it). Instaed, they could just invest in this equity, as it is not burning the HYPE it buys, but is increasing its share of it. so instead of buyback-n-burn, its closer to commit-n-compound. I’ve started thinking about this as a value-accrual architecture, with two different forms of capital (terms best described by a16z crypto's post in 2025): Network asset: scarcity at the core. - Its supply discipline can become more credible, its economics more programmatic, and its operation less dependent on any single company. - It coordinates shared infrastructure and connects activity across many businesses to a common asset. - That is what I mean by a progressively hardening asset. The scarcity and independence have to be earned through the design of the system. - Not capital intensive, as it must be smart contract based. - Can often place be the seat of the "decentralized security" - as there is no central underwriter. If HYPE was a company.. itd be regulated out of the wazoo. Enterprise equity: elasticity at the edge. - A company can issue shares, raise capital, hire people, acquire licenses, and build products. - Its capital structure can expand with the work required to bring the network into the real world. - New shares dilute ownership percentages, but productive investment can increase value per share. - That ability to finance expansion is why I think of equity as elastic - This is the part where its OK to be regulated, because most of the regulational burden is held by HYPE, theyre just asset issuers. These different traits are the reason to combine them. An enterprise can raise the capital to grow while the network maintains its own supply discipline. The connection is the activity that expansion creates, and the mechanisms that route value from that activity back to the network. Hyperliquid makes part of that relationship explicit. Trading generates fees for the deployer and the protocol. The Assistance Fund (insurance) automatically converts its allocated fees into HYPE, which the documentation specifies is burned. Hyperliquid fee mechanics The broader strategy I take from this is Commit & Compound that is conducted around the equity enterprise. It's effectively what a DAT 2.0 would do: create a business first, then max-bid HYPE with that entity, then get new investors to invest in the convexity of HYPE^2. Build a useful business around the network. Commit capital to its asset. Retain exposure to the infrastructure your business helps make more valuable. For a company that accumulates and holds the network asset, growing enterprise value and participating in network value become connected opportunities. There is a business to capitalize and a shared ecosystem in which to hold an economic stake. That is why I find the assumption that every successful application eventually needs its own token less compelling. Building a valuable company around an existing network asset is a meaningful strategy in its own right. Notice how much kinetic language I’m using: commitment, circulation, gravity I think network assets and enterprise assets represent different forms of gravity HYPE is the sun: the network asset at the center tradexyz equity is a planet: an enterprise building its business around that network Its real-world integrations are moons: the licenses, institutional relationships, and distribution channels that gather around the enterprise and extend its reach Enterprise value forms around network value. Real-world relationships form around enterprise value. Their activity feeds back into the shared system. Personally, I think the greatest value can ultimately accrue at the network level. But that belief has to be substantiated by the businesses being built around it: real users, useful products, liquidity, and recurring economic activity. A token doesn’t acquire gravity simply because we call it a network asset. This is the distinction I found useful in a16z’s writing: can the network function and accrue value independently of the original company? Can other businesses build on it and contribute to its economics? The company can become a great business. The network can become the foundation for many great businesses. Scarcity at the core. Productive expansion around it. I’ve learned a lot from this Commit & Compound structure, with a Copernican asset at the center. It has changed how I’m thinking about USDAI and CHIP. More on that in the next post

Compute capital markets subsumes “agentic commerce” because it establishes the economic hurdle rate against which all agentic activity must be measured Compute is a cost of capital, or in other words, an interest rate @USDai_Official

A year ago since we launched we got a ton of criticisms about how we worked, but that’s fine. It’s called v1 for a reason With the launch of GLX, @USDai_Official has effectively undergone a “Ship of Theseus” and evolved every pillar of the protocol: redemption, liquidity, mints

lots of things to unpack: 1. our largest deal yet (big size) 2. our 1st investment-grade offtake deal (great credit) 3. a new product offering to help scale our business - that k3 annc was a part of it you think you know what USDai is, but the design is improving each month

The objective of @USDai_Official is to create a financial platform for the next 10,000 neoclouds The GPU mortgage is just the start, afterwards it’s settlement, and then it’s the entire financial stack as all dollars in AI does one thing only: pays down the mortgage

The answer is that ai companies need debt access directly Unfortunately DCM is not suited for GPU mortgages @USDai_Official enables the GPU mortgage and compute ownership to be possible at scale. Skip the Goldman desk, skip the hyperscaler rent, and get the CHIP directly

Series A: secured 💰 GPUs: on the way 🔥 Data centers: deploying... We need a Senior Network Infrastructure Engineer to wire it all together. Cloud, on-prem, Kubernetes, security. Come build the backbone of private AI. Join @AskVenice

Databricks CEO: "We host open-source models such as Kimi and offer them to our customers. Demand has been so strong that we are running out of GPUs across multiple regions. We nearly exhausted our GPU capacity in Asia, and demand is rising in countries including Japan, South Korea, the United States, and India. We, therefore, need to acquire a large number of additional GPUs, which requires significant funding. That demand was what triggered our latest fundraising round: we were inundated with customer requests and needed more GPU capacity. GPUs are extremely expensive to acquire.

David Choi (@0xZergs ) is speaking at Money Frontier 2026. David Choi is the Co-Founder and CEO of Permian Labs, the developer of @USDai_Official . With a background in structured finance at Deutsche Bank and as an early DeFi investor and MEV participant, he is building onchain capital markets for AI infrastructure. Join him at Money Frontier 2026. Get your tickets and learn more:

How @USDai_Official generates the highest stablecoin yield above $100M, per founder @0xZergs: - Deposit stablecoins - Capital gets lent against GPU chips in real data centers - "The highest yield for any stable above $100M...no one else is close" - Every loan visible on the live world map on their website - Utilization heartbeat and an underwriting matrix for scoring assets coming soon $usdai | $CHIP