Accretion depends on the metric. Net Assets per Share measures balance sheet strength and residual asset value. BTC per Share measures Bitcoin intensity and long-term equity upside. NAV accretion improves asset coverage. BTC Yield accretion increases Bitcoin per share. $MSTR $BTC

Where Does Value Accrue in the Agent Economy? Visa's Head of Crypto on stablecoins, x402/MPP, and card networks Shoal Signal with @cuysheffield, Head of Crypto at @Visa, hosted by @zaddycoin We cover where value accrues in the agent economy, stablecoins vs. cards and the value propositions, the trust problem with open x402 endpoints, autonomous agents, headless merchants, why agents buying tokens will dwarf everything else, and why he thinks agents should stay accountable to a human. 0:00 Intro 1:49 Where does the value accrue 2:31 Payment networks vs payment infrastructure 5:24 Trust and governance is the missing piece 6:37 Anyone can spin up an x402 endpoint 7:53 Why a crypto vet won't give agents a wallet 10:17 Headless merchants are the big opportunity 11:35 Sell workflows, not tools 13:43 The demand side is the harder problem 19:25 Where the first billion gets spent 23:15 Token share on inference 25:02 Problems with agent spend 26:49 Agents buying tokens dwarfs everything 33:04 Should agents be self-sovereign? 35:06 Every individual becomes their own corporation 36:24 Visa Intelligent Commerce and TAP 39:22 Visa CLI for developers 41:30 The browser-to-checkout gap 44:31 New merchants built for agents 46:28 Where the alpha is

A framework to understand how value accrues across the AI stack. This is a blueprint for understanding what builds AI into its pragmatic parts: what each layer is, where it ends, and where value is accrued. So here’s how you can think about it: 1. Layer 1 - Infrastructure Before any AI model trains or any robot moves, an industrial foundation must exist. Land, energy grids, cooling systems, critical minerals, and fabrication facilities. Infrastructure is the constraint that all the other layers depend on. 2. Layer 2 - Chips Transistors that are etched onto silicon wafers using extreme ultraviolet light. This is what allows both physical and digital AI to take an input, process it, and return a predictive output. The more transistors that fit on a chip, the more computation it can perform. 3. Layer 3 - Data Both digital and physical models train on data. Digital models train on text, code, and images; physical models train on gravity, friction, depth, and sensor streams. The more accurate the data, the more accurate the output. 4. Layer 4 - Models A model is a system that learns from examples. Feed it enough examples of inputs paired with correct outputs, and it adjusts its internal structure until it can predict correct outputs on inputs it has never seen before. LLMs represent a specific class trained on text. They learn by processing billions of examples of human language, developing the ability to write, reason, summarize, and generate code. 5. Layer 5 - Execution This is what lets models take actions on behalf of users. The execution layer lets models pursue objectives through sequential action: observing the environment, reasoning about the next step, acting, and looping until the goal is reached. 6. Layer 6 - Application All of the AI Stack’s revenue originates at the application layer, then goes to the layers below. Every dollar paid for AI is paid for an outcome, a task completed, and an answer delivered. Nobody wants H100s for their own sake. They want H100s because someone, somewhere, wants to run an application. These are the different layers that make up the entire ecosystem of AI. We did a full study on the AI stack. If you want to read about it, head over to my Substack (

Ukraine’s Ministry of Defense-accredited WeTrueGun drone school has turned Grand Theft Auto V into a practical FPV drone simulator. Pilots connect real radio controllers such as the Radiomaster TX16S and train in the game’s vast open-world map, which includes thermal vision and dynamic moving targets. The mod takes advantage of unrestricted urban environments, vehicles, and pedestrians to closely replicate actual combat conditions without the boundaries or scripted scenarios found in dedicated simulators. It serves as a low-cost supplement for sharpening skills, experimenting with mission designs, and keeping operators sharp between live training sessions. The school’s full 35-to-37-day program already covers drone assembly, tactics, electronic warfare countermeasures, situational awareness systems, and more.