Happy Heavenly Birthday Rick Wright! 🎹 Born July 28, 1943. Co-founder of Pink Floyd, his atmospheric keyboards and songwriting helped define the band's timeless sound. A Rock & Roll Hall of Fame inductee, his musical legacy continues to inspire generations. Forever remembered🎶

Wendy’s is bringing back a familiar look that longtime customers have been asking to see again for years. Starting Sept. 28, the fast-food chain will roll out its iconic yellow packaging for burgers, fries, nuggets and Frostys, reviving a look that was a staple at Wendy’s for decades. “Fans asked. Constantly,” Wendy’s marketing chief says, calling it one of the brand’s most requested comebacks. The throwback packaging will stick around into 2027 as Wendy’s joins other major chains tapping into nostalgia for their old-school look.

Everyone owning $MU is trying to answer one question: Do record memory prices mean the cycle is about to roll over? (Save this) The case for no starts with NVDA evaluating 33% less HBM for its next architecture. They would not redesign around a part that is abundant! Next, Memory spend currently at ~33% of total AI data center spend is expect to head to ~50% next year. Also, the Moonshots clip below explains why supply-side is paranoid about oversupply from past boom-bust cycles. Which is why they are likely to not overbuild and also sign LTAs with floor prices. However, there is one scenario (among others) where the memory trade could eventually break. Today, AI repeatedly pulls model weights from memory to generate answers. If those weights become stable enough, you could bake them directly into the chip and remove much of the memory needed. That works best for mature, repetitive workloads where the same model can run for a long time. The tradeoff is flexibility. Once the model is baked into silicon, you cannot simply load the next version onto the same chip (=stranded assets). My view is that 2026-28 is about using less HBM for each job, not eliminating memory. Beyond that, innovation like the described case above could become materials. At Milk Road PRO we have been banging the drum on memory since launching our portfolios in March. We bought the dip recently but also sold some of our AI infra positions. All with live notifications! Come join us before prices increase after Aug 26:

What is Stellar’s “Adapter” Upgrade? Stellar’s (@StellarOrg) next network upgrade, Protocol 28, is called Adapter. Stellar describes it as a builder-focused upgrade, with most of its changes aimed at making Soroban smart contracts easier to build, upgrade and maintain. But Adapter also includes a deeper change to Stellar’s consensus process that could help the network handle growing activity more efficiently. Here’s what is changing: (1) Faster consensus under heavy network load CAP-83 changes how Stellar validators handle transaction data during consensus. Today, validators need to receive a full transaction set before they can make progress. Adapter allows validators to begin voting before the entire transaction set has arrived. If the transaction data is late or invalid, validators can explicitly drop it instead of waiting indefinitely. That means consensus can keep moving even when transaction data is slow to propagate. The performance gains will roll out gradually as parallel transaction set downloading is enabled after mainnet. For users, this is mostly a behind-the-scenes improvement. For Stellar, it could become important as the network scales and needs to maintain fast, low-cost transactions under heavier demand. (2) Smart contract fleets can be upgraded together CAP-85 introduces externally managed executables for Soroban contracts. This targets protocols that operate large fleets of contracts using the same underlying code. Today, upgrading those contracts individually can leave some instances running old code while others use the new version. That creates both operational complexity and a potential security risk. With Adapter, contracts can point to a shared executable reference controlled by another contract. Developers can update that shared reference once, allowing every connected contract to switch to the new code atomically. (3) Smart contract data becomes easier to migrate CAP-86 tackles another problem developers face as contracts evolve. Smart contract data structures can change over time as developers add fields, remove old ones or modify interfaces. Existing host functions can reject data when its structure does not exactly match what a contract expects. Adapter introduces new sparse host functions that can handle missing or additional fields more gracefully. That makes it easier for developers to migrate contract data between different versions without breaking existing applications. For builders, this could make long-lived Soroban applications easier to maintain. (4) Adapter is mainly about developers The important point is that Protocol 28 is not primarily a user-facing feature upgrade. Two of its three major changes, CAP-85 and CAP-86, are focused directly on smart contract development. They aim to make contracts easier to upgrade, safer to maintain and more flexible as applications evolve. CAP-83 works at a deeper infrastructure level by improving how Stellar reaches consensus when transaction data is delayed. Together, the changes are aimed at making Stellar more capable as its smart contract ecosystem grows. So when does Adapter actually arrive? Stellar has scheduled the Protocol 28 testnet upgrade vote for August 27 at 17:00 UTC. The mainnet upgrade vote is scheduled for September 16 at 17:00 UTC. The mainnet vote is the key governance decision. A successful vote would move Protocol 28 closer to deployment, but the vote itself does not mean Adapter is already live. Stellar also expects its Core stable release to become available on August 13, followed by additional infrastructure and SDK releases between August 13 and 21. Developers, validators and infrastructure operators will need to prepare for the upgrade before the relevant votes. If Protocol 28 passes its mainnet vote, Stellar will take another step toward becoming a more mature platform for large, long-lived smart contract applications.

Grok Bot summary of SpaceX CFO Bret Johnsen’s new interview with Tony Kim at SpaceX Mission Control. Bret’s path - Came from semiconductors (Qualcomm, Broadcom; was CFO of MindSpeed) - Barely knew SpaceX / Elon when a board member called - Thought aerospace sounded “old and stodgy” vs 6-month chip cycles - Visited the rocket factory, called his wife on the drive home: we’re moving back to LA - Now sits in the same magical factory: Dragon in the rafters, ~700 launches run from Mission Control, ~40 buildings feeding the campus - Full circle: semiconductors again via AI / chips --- Taking on AI complexity (during the biggest IPO ever) - SpaceX DNA: run toward “impossible” challenges (first booster recovery felt like a rock concert) - xAI acquisition in January + X + Cursor + Colossus = complexity multi-X in ~1–2 years - AI is now the biggest business at SpaceX - Bought time-to-market into the biggest product cycle of their lifetime - SpaceX leadership DNA applied to that team - Already feel leading on infra cost and speed - Products moving: Grok 4.6, 4.7 soon, Grok Bot viral, Cursor strong - Every week matters; soon every day --- AI business: the money picture - ~$100B company ARR exiting this year (annualized on December) — AI is more than half / the majority - A year ago AI revenue was ~zero - Capacity: a bit over ~2 GW this year → guidance 5–10 GW by end of next year - Becoming one of the biggest hyperscalers, fast Why they move so fast (vertical integration) - Engineer the data centers themselves - Act as GC: buildings, power, cooling, networking (plus partners) - Used to regulators from rockets + telecom - Removes margin stacking - Talent transfer: launch-site / pressurized systems people onto data centers — Elon compared it to Yankees vs little league Hosting model / ROI - Monetize compute while catching up on software - Demand knocked: “what are you doing with all that Colossus compute?” - Short contracts: ~3 months + 3-month out (flexibility for their own future training needs) - Payback: dramatically less than one year on cutting-edge terrestrial compute - Hosting customers include tier ones like Anthropic, Google, and others - Same watts can later flip to their own software at higher $/watt — optionality Frontier models - Moving maybe faster than anyone on trajectory - ~5 serious players; need to be at frontier for recursive improvement - Coding solutions improving dramatically - Consumer vs enterprise tools will stratify - Grok Bot: growing faster than anything they’ve seen; “transformational” user feedback; enterprise coming; expects roll into Cursor AI safety - Follow Elon’s self-regulation / peer-review path (like other industries) - Logical, collaborative, including outside the US - Hopeful something lands quickly TAM framing (IPO) - ~$28T TAM quoted at IPO; AI was ~$25–26T of that (digital AI only) - Physical AI not even in that number — Bret thinks physical AI eventually bigger than digital - Launch / space TAM = hundreds of billions (almost rounding vs AI) - Connectivity (mobile + broadband) ~$2T — still small vs AI - Differentiation: infra + X content (~500–600M MAUs) + consumer + enterprise push (Cursor etc.) What to watch: Starship + terrestrial compute Bret’s prioritization cheat code: 1. Get Starship right 2. Get terrestrial compute right for the next ~2 years until orbital kicks in → Transformational financially Starship status - Catalyst for all three businesses: space, satellites, AI - Holy grail: full reusability → another ~10x cheaper $/kg to orbit - Opens moon boots / lunar base / bigger industry - Flight 14 later this month: production V3 Starlink broadband satellites — real revenue payload - One Starship V3 load ≈ ~20× what Falcon 9 flew for Starlink capability - Each V3 satellite ~1 terabit; ~50–60 per flight = enormous bandwidth per launch - A couple launches start filling V3 planes; first-half-next-year benefits - Eventually Starlink as % of global internet traffic - Connectivity demand from machines (robots, AVs, agents) will dwarf human brains - Same Starship unlocks next-gen direct-to-cell / direct-to-device - Same bus + more solar + radiator + compute chips = orbital compute constellation Reusability timeline - Flight 15 may attempt catch / bring back first + second stage - First stage already recovered and reflown - Flight 13 soft splash best yet; recovered ship like a whale, engineers studied heat shield - Reliable stage recovery this year; meaningful reflight next year - Once a few second stages reflight → orbital compute cost can match terrestrial - Then orbital cost curve goes down (tech / Moore’s law / aircraft-like amortization) while terrestrial gets more expensive (power, labor, regulation, cooling) - Clean free solar, no community impact - Industry agrees space compute is the future — debate is timing; others say ~decade; SpaceX flies first ones next year; meaningful compute in ~2028 Scale ambition - Both cost and speed matter; near-term availability/speed most important - On track toward ~100 GW/year by end of decade (~20% of entire US grid each year) - Elon talking terawatt-scale in the 2030s - Hosting proves staying power + virtuous cycle for industry - New NVIDIA platforms: just fly the next satellite generation — no whole new terrestrial campus cycle TerraFab (chips / memory) - Parallel to orbital ramp, not after - Not replacing NVIDIA — love the partnership; will buy all they can - Worry is deeper fab / memory supply chain - Memory industry ~30%/yr growth vs SpaceX wanting triple-digit — won’t work - Hard technical challenge; memory may be early focus - Want volume benefit by late decade (demo next year with Tesla development line → something in ’28 → meaningful ’29+) Viewer Q&A highlights **First bottleneck in the AI chain?** - Chip supply for next year - Power for infra players (utilities used to ~1% growth; turbines / blades constrained) - SpaceX standing up investment casting for turbine blades themselves - Orbital is the big unlock; squeeze every terrestrial watt meanwhile How to tell durable AI spend vs keep-up spend? - Bret: durability of revenue; diversified monetization of the same compute; host now, software later at higher $/watt; guided next-year monetization range ~$30–50/watt - Tony: stack layers have different costs, margins, contract duration; SpaceX spans many layers with <1-year payback on infra How soon a data center in space? - First orbital compute satellites end of next year - Meaningful compute in 2028 - Better phrase: “racks in space,” not giant buildings - Same V3-like bus + solar + radiator + compute chips - Already flown ~11,000 sats; launch + inter-sat + ground links already solved - ~full rack (~72 NVIDIA-class) concept per sat; ~5 MW compute per launch Tony’s investor wrap - Still early; trends accelerate into physical AI and beyond Earth - Connectivity / tokens increasingly for machines, not just humans --- Bottom line in one breath SpaceX is using rocket DNA — vertical integration, crazy talent, short-cycle iteration — to become a top AI infra + software player now, while Starship unlocks Starlink V3, direct-to-cell, and orbital “racks in space.” Watch two things: terrestrial compute scale (toward $100B ARR and multi-GW) and Starship reusability. Everything else is downstream of those.