💰 Don't size caches for peak—let them stretch and shrink with demand to cut cost. Google applies the classic ski rental problem to a production database. Title: Optimizing cloud economics with linear elastic caching URL: 📦 Overview Linear elastic caching treats memory footprint as a variable cost that integrates over time, dynamically growing and shrinking cache size to match the workload. 🎯 The problem Cloud memory is expensive (serverless can charge up to $3/day per GiB). Fixed-size caches hit a "Goldilocks dilemma": too small hurts performance, too large wastes thousands on idle memory during low demand. 🎿 Method Each page faces a choice: "rent" (keep in RAM, paying continuous memory cost) or "buy the miss" (evict, risking latency/I/O later). A ski rental algorithm sets each page's TTL. The key result: eviction policy and rental duration can be optimized separately. 🌲 Implementation For Spanner, a lightweight shallow decision tree (compilable to C++) predicts the optimal TTL from data size, miss cost, and operation type—no heavy inference in the cache path. 📉 Results In production on Spanner: memory down 15.5%, misses up only 5.5%, TCO down ~5%, I/O impact a mere 0.5%. On public traces it consistently beat fixed-size (GDSF) baselines. #CloudComputing# #Algorithms#

one funny thing agents are exposing is how many businesses are just built on human laziness they count on you being too lazy to cancel, appeal, claim, or fill out the paperwork, which worked great for them until AI agents showed up things you should send your agent after right now: > cancel subscriptions you forgot about (~50% of subscription revenue comes from people forgetting to cancel) > earn 30-50x more by moving your idle cash into high-yield savings (checking pays 0.1% on average, high-yield pays 3-5%) > negotiate your internet and phone bills (1 Muse user had it negotiate with Comcast and saved $800+/year) > claim unclaimed money in your name (states are holding $70B+, owed to 1 in 7 Americans) > track down old 401ks ($2.1T is sitting in forgotten accounts) > spend your health spending account (FSA) before it expires (47% of holders forfeit $422 on average) > appeal insurance denials (67% of appealed Medicare Advantage denials get reversed) > file for class action payouts you qualify for (a record $79B in settlements last year, and only ~9% of eligible people claim)

Two halves of one problem. What you feed the model, and what you do with what comes out: - Krea threw AI images out of the training set on purpose, because distilled data is sticky and the shortcut costs you the whole point. Roughly 30 to 40 in house classifiers run over billions of images - GPU utilization is a lie. It reads 100 percent while the tensor cores sit idle. Any GPU running above 78 degrees gets pulled rather than debugged - 1.3 million videos rendered by open source users in 90 days, from a framework deliberately designed so a small model could author it - An avatar generating continuously for eight hours with no reset, and a sixteen hour deployment being built. It costs about what serving a voice model costs - 10 dollars now buys three hours of continuously generated video. The old way was a slot machine at roughly 10 dollars a minute Watch the Generative Media track:

Circle (@circle) launched cirBTC on Ethereum on June 8, a 1:1 BTC-backed ERC-20 where each unit sits in segregated, regulated custody at a Circle entity, marking a new product line for Circle beyond the stablecoin categories. The tokenized Bitcoin market currently exceeds $10 billion in total supply across all wrapped formats, one of the most actively used collateral asset classes in DeFi, yet still represents under 2% of Bitcoin's total market cap. The vast majority of Bitcoin value sits idle, never deployed as collateral or generating yield. For any issuer that can solve the custody and trust problem convincingly enough to attract that capital, the addressable market is orders of magnitude larger than what exists today. WBTC and @coinbase cbBTC are the two dominant wrapped Bitcoin formats on Ethereum, together representing the majority of tokenized Bitcoin supply deployed in DeFi. The protocol that makes most of that supply productive in DeFi is @aave, being the primary collateral venue where wrapped BTC is usded as the basis for borrowing, leverage, and liquidity strategies. Looking at the top 50 Ethereum holders of WBTC ($5B) and cbBTC ($2.5B), Aave V3 absorbs $3.1B (41% of combined supply), more than Morpho Blue, L2 bridges, Compound, SparkLend, and all DEXes combined. There is already an ARFC to onboard cirBTC on both Aave V3 Core and Aave V4 Core on Aave governance, also bringing DeFi utility to Circle new asset. Running in parallel on Aave V4, @babylonlabs is integrating Trustless Bitcoin Vaults into Aave V4. Under this model, users lock native BTC directly on the Bitcoin blockchain in Taproot UTXOs and use that position as collateral to borrow stablecoins on Aave V4 through a dedicated Babylon Core Lending Spoke, without wrapping, bridging, or giving up custody at any point. Babylon has already secured 51k BTC trustlessly. This is the first major native BTC collateral primitive on Aave V4. Bitcoin's market cap stands at $1.32 trillion. The entire tokenized BTC market represents roughly +$10 billion in deployed supply, under 2% of that total. Every incremental unit of Bitcoin that moves into regulated wrappers like cirBTC, or gets deployed trustlessly through primitives like Babylon's native BTC vaults, expands the total productive capacity of DeFi in a way that hardly any new stablecoin or synthetic asset can replicate.

I am the Vice President of Ad Integrity at Meta. I want to talk about the number sixteen. Sixteen billion dollars. That is what we earned from advertisements our own internal classification system flagged as "higher legal risk." Crypto scams. Romance fraud. Impersonation schemes targeting the elderly. We had a dashboard. The dashboard had a color. The color was green. Green meant revenue. Three point five billion every six months. I watched that number on the Revenue Integrity Dashboard every Monday at 9 AM. The same meeting where we reviewed takedown requests. The same room. We did not remove the ads. We removed 8,000 people. The memo said "efficiency." The memo said "leaner teams." The memo said "AI-first." What the memo did not say: the 8,000 people we fired cost us $4.2 billion annually in compensation. The ads we refused to remove earned us $16 billion in the same period. The math was never complicated. The math was the strategy. I received the Ad Quality Excellence Award in 2024. It is on my desk. It is a glass rectangle. It weighs more than the compliance reports we filed with the FTC claiming we had "robust systems" to prevent fraud. But I want to talk about April. In April, we installed software on every employee laptop in Building 20. The software tracks mouse movements. Keystroke cadence. Application switching. Idle time. It sends a report every eleven minutes. We call it a "productivity signal." The advertisers call their version "behavioral data." Same architecture. Same team built both. I know because I approved the vendor contract for the external version in 2021 and the internal version last month. The vendor is the same. The codebase is the same. The only difference is the target. When we track users, it's a $140 billion business. When we track employees, it's "performance management." When the employees objected — posted in the internal channel, filed concerns with HR, asked the obvious questions — we did what we always do. We reminded them of the NDA. We reminded them of the stock vesting schedule. We reminded them that 8,000 people were no longer receiving reminders of anything. They stopped posting in the channel. I am told the keystroke heat map is displayed on monitors in Building 20. I am told it updates in real time. I am told it looks exactly like the user engagement dashboard we show advertisers. I am told this is a coincidence. The product has always been the person. The only variable is which person. For sixteen years, it was the user. Their clicks. Their attention. Their data. For the advertisers, it was their money. Clean or dirty. We did not ask. Asking would have cost us $3.5 billion every six months. Now it is the employee. Their keystrokes. Their idle seconds. Their bathroom breaks quantified as "disengagement intervals." We are a platform that earned $16 billion from fraud we refused to stop, fired 8,000 people to "cut costs," and now tracks the survivors' mouse movements every eleven minutes to ensure they are sufficiently productive. The product is the person. The person is the product. That's the platform.

Over the past month, there has been no shortage of headlines about struggling exchanges. Some of the founders involved are people I’ve crossed paths with. This is a sobering reminder of how unforgiving every crypto downturn can be. Most conversations today focus on which exchange survives or which project fails. But after spending the past month in the US and Europe meeting with Wall Street traders and institutional clients, I came away with a different perspective. What surprised me is that the institutions many people see as the backbone of market liquidity aren’t exactly having an easy time either. Many are going through painful adjustments of their own. A few observations that stayed with me: 1. The scars from last year’s 10.10 market shock are still healing The decline in crypto liquidity wasn’t temporary, but structural. Comparing notes with several institutional trading firms, even the largest venues have generally seen liquidity decline by around 30-40% since the market turmoil. Lower liquidity isn’t just about lower trading volumes. It also raises the cost of trust. Higher volatility, more fragmented liquidity and greater sensitivity to market manipulation have made institutions much more cautious about deploying capital. One lesson has become very clear: don’t overestimate how quickly markets recover, and don’t underestimate how long it takes to rebuild confidence. Everyone is repairing balance sheets. That process takes time. 2. Long-term conviction hasn’t disappeared, but the playbook has changed Despite a difficult market, institutional interest in digital assets hasn’t gone away. If anything, more firms are quietly preparing for the next cycle while prices remain subdued. The biggest change is how they think about crypto. It’s no longer viewed as a standalone speculative asset class. It’s increasingly becoming one component within a broader global portfolio. Multi-asset strategies, tokenized real-world assets, cross-asset collateral and hedged portfolios are becoming standard discussions. That also helps explain why some of the crypto trading volume lost over the past year is being replaced by equities, FX and commodities. Institutions aren’t leaving. They’re evolving. They’re optimizing for more stable, diversified return profiles rather than relying on pure crypto beta. Platforms built only around crypto trading may find it increasingly difficult to meet those changing needs. 3. More than ever, institutions want peace of mind From FTX to the more recent incidents across the industry, every exchange crisis has reinforced the same lesson: safety is the minimum requirement for staying at the table. When I speak with institutions and VIP clients today, the conversation is no longer just about generating alpha. Asset security, risk management and capital efficiency now matter just as much. They don’t want to put all their eggs in one basket. At the same time, they don’t want their capital sitting idle or becoming fragmented across different platforms and accounts. What they are looking for is fairly straightforward: transparent third-party custody, clear risk controls and an account structure that allows capital to move flexibly when opportunities arise. This is also why products such as rToken are attracting more attention from professional investors. The same position can provide market exposure, be pledged to access liquidity and be used as margin. The goal is not to take more risk with the same capital. It is to make every dollar work harder while keeping safety at the centre of the equation. ------- The financial industry has always rewarded scale and trust, and crypto is no different. I’ve often told our team that many offshore exchanges outside the top 10 group may not survive the next few years. But even being among the largest players is no reason to become complacent. Bear markets are uncomfortable, but they have a way of forcing everyone back to fundamentals. The companies that emerge stronger won’t simply be the ones that cut costs or survive another cycle. They’ll be the ones that manage short-term risk while continuing to build infrastructure, discover genuine product-market fit and solve real customer problems. That’s what we’re focused on. And I believe that’s where the industry’s next chapter will be written.

$LIN KEY READ-THROUGHS FROM LINDE’S Q3 2026 EARNINGS CALL Linde’s October 29, 2026 call reinforces a selective industrial expansion rather than a synchronized global recovery. The strongest positive signals concern U.S. refining, semiconductor manufacturing and data-center construction. The clearest negatives concern European industrial competitiveness and the timing of returns from integrated low-carbon projects. Pricing requires particular care: higher gas revenue can represent recovery of energy and logistics costs rather than incremental profit. Several affected companies have already reported Q3 results, making Linde’s commentary corroborating evidence for subsequent quarters rather than a fresh prediction of their reported earnings. Impact assessments distinguish the importance of the underlying operating exposure from the incremental information supplied by this call. Countries identify corporate headquarters; magnitude assessments are fundamental, not share-price forecasts. ENERGY AND DECARBONIZATION LOW-CARBON AMMONIA: COMMISSIONING AND CASH RETURNS REMAIN SEPARATE MILESTONES (READ-THROUGH 1) Affected company: Woodside Energy Group (WDS: Australia). Direction: negative for the timing and certainty of Beaumont’s investment returns. Magnitude: high at the project level, but low to moderate at the consolidated group level. Call evidence: Linde has started nitrogen supply and is temporarily supplying hydrogen from its existing Gulf Coast network. Its 2 dedicated autothermal reformer trains are expected to complete their staggered start-up in 1H27. Carbon capture and storage remains dependent on ExxonMobil’s infrastructure. Lamba indicated that the more meaningful contribution would emerge in 2H27, with Linde’s facilities potentially reaching full utilization in 2028. The transmission mechanism is delayed conversion of invested capital into fully integrated, lower-carbon production and cash flow. Nitrogen availability, hydrogen commissioning, carbon-storage readiness and customer utilization are distinct milestones. Completing an individual component does not establish commercial readiness of the entire system. Woodside can already produce conventional ammonia using temporary feedstock supply, so the call should not be interpreted as implying no ammonia production before 2027 or the removal of the plant’s entire capacity from the market. The more subtle issue is the opportunity cost of interim hydrogen supply. White stated that these molecules could otherwise have been sold elsewhere in Linde’s network. This confirms that temporary supply is drawing on commercially valuable capacity rather than economically idle infrastructure. It does not establish that Woodside pays spot hydrogen prices, bears all those costs or owes additional contractual penalties. It does establish why dedicated supply is important to improving the integrated project’s economics. Woodside’s October 21 report had already disclosed completion of its strategic review and an intention to begin a formal process to reduce its investment in Beaumont. Consequently, Linde’s call is not the first disclosure of potential ownership change. Its incremental value is the supplier-level explanation of commissioning dependencies and the extended utilization ramp. Woodside’s reported feedstock constraints and approximately 30-day operating suspension further demonstrate that plant ownership, feedstock supply and commercial delivery are separate execution risks. Near-term catalysts are the terms of any investment reduction, evidence of reliable feedstock supply and progress toward hydrogen commissioning. The longer-duration issue is whether lower-carbon production generates sufficient incremental economics to compensate for capital tied up during the extended ramp. The latest schedule was reaffirmed, not newly delayed on this call; the negative read-through concerns the return profile and remaining execution burden rather than a newly announced guidance cut. COMPLEX REFINING: HYDROGEN DEMAND VALIDATES FEEDSTOCK FLEXIBILITY AND DISTILLATE EXPOSURE (READ-THROUGH 2) Affected companies: Marathon Petroleum (MPC: United States) and Valero Energy (VLO: United States). Direction: positive for operating conditions at complex refineries. Magnitude: potentially high within refining earnings, but moderate incremental information from Linde because strong industry conditions are already visible elsewhere. Call evidence: Chemicals and energy sales increased 12%. Management identified stronger U.S. Gulf Coast activity and specifically highlighted “more complex refiners requiring more hydrogen to process heavier and higher sulfur crude.” White also emphasized hydrogen’s role in producing middle distillates and ultra-low-sulfur diesel. The precise transmission mechanism is refinery complexity, not simply higher crude prices. Hydrogen enables desulfurization and upgrading of difficult feedstocks. Refineries capable of processing heavier or higher-sulfur crude can benefit when those feedstocks are discounted relative to the higher-value products they produce. Strong hydrogen demand provides physical operating evidence that this upgrading capacity is being used. Marathon Petroleum and Valero have substantial complex refining exposure, including Gulf Coast operations. The read-through favors assets with feedstock flexibility and sufficient conversion capacity over a generic upstream-energy basket. Higher oil prices alone do not guarantee stronger refining earnings; the relevant variables are feedstock differentials, product margins, utilization and operating reliability. The qualification is that hydrogen demand does not reveal realized refining margins. More intensive crude processing also consumes additional hydrogen and energy. Linde’s commentary supports throughput and processing-severity assumptions, but does not establish the net product-minus-feedstock spread or prove an earnings beat for either refiner. Marathon Petroleum’s November 3 results provide the nearer trading catalyst, particularly commentary on utilization, maintenance, crude selection and distillate capture. Valero’s October 22 reporting date precedes Linde’s call, so the signal is principally corroboration and support for subsequent operating assumptions. Longer term, complex refining assets retain an advantage when feedstock quality and trade routes are disrupted, but exceptional conflict-related margins should not be capitalized indefinitely. INDUSTRIAL GASES AND SEMICONDUCTOR MATERIALS INDUSTRIAL HYDROGEN: POSITIVE PHYSICAL DEMAND FOR AIR PRODUCTS, WITHOUT A PASS-THROUGH WINDFALL (READ-THROUGH 3) Affected company: Air Products and Chemicals (APD: United States). Direction: positive for the Americas hydrogen and synthesis-gas operating environment. Magnitude: moderate at the relevant business level and low to moderate for consolidated earnings. Call evidence: Linde described U.S. merchant and on-site volumes as its strongest, highlighted Gulf Coast chemicals and refining demand, and characterized hydrogen as a valuable molecule in the current market. Its temporary supply to Woodside is consuming network capacity that management believes could otherwise serve additional customers. Air Products has a directly relevant operating exposure. Its preceding fiscal quarter attributed Americas volume growth partly to hydrogen and carbon-monoxide facilities and a new on-site asset. Linde’s September-quarter commentary therefore provides a later, independent indication that the regional demand environment remains supportive. This is a shared-market read-through, not evidence that Air Products supplies Woodside or has won any Linde customer. The transmission mechanism is customer offtake and utilization of existing production and distribution infrastructure. Stronger physical demand can improve the economics of available network capacity and support incremental sales where contracts permit. Long-term on-site arrangements also provide more revenue protection than an uncontracted commodity producer possesses. However, energy inflation should not be counted as equivalent profit growth. Linde stated that on-site energy adjustments are generally recovered within the current month, while merchant and packaged recovery can lag approximately 3-6 months. Air Products’ own prior release identified distribution and dislocation costs as pressures. Higher reported sales can therefore coexist with modest profit improvement or percentage-margin dilution. The November 5 fiscal Q4 report is the immediate catalyst. The useful tests are Americas volume growth, hydrogen-network demand and profit conversion after separating energy pass-through and accounting effects. Linde’s margin performance should not be transferred mechanically to Air Products, whose preceding results also benefited from lower depreciation on certain assets classified as held for sale. The longer-duration positive is demand for established industrial hydrogen infrastructure serving existing customers. That is economically distinct from speculative demand for new low-carbon hydrogen projects and should be valued separately from those projects’ commissioning and commercialization risks. SEMICONDUCTORS: GAS CONSUMPTION CONFIRMS OPERATING DEMAND, NOT JUST ANNOUNCED FAB SPENDING (READ-THROUGH 4) Affected companies: Air Liquide (AI: France) and Taiwan Semiconductor Manufacturing Company (TSM: Taiwan). Direction: positive. Magnitude: high within semiconductor-gas activities, moderate for Air Liquide’s consolidated growth, and a modest incremental operating confirmation for TSMC. Call evidence: Linde’s electronics sales increased 15%, with management describing growth as largely organic. Its 50%-owned Taiwan joint venture, excluded from that consolidated growth figure, was growing at a similar rate. Approximately $2 billion of electronics projects are under execution, with staggered start-ups over the next 2-3 years. Management explicitly identified ongoing discussions with TSMC in the U.S. and emphasized its preference for logic-related projects. The important distinction is between operating consumption and prospective capacity. Organic gas demand indicates activity at existing semiconductor facilities, while the project backlog supports future installed capacity. The combination is stronger evidence than backlog growth alone because it points to both current production requirements and future expansion. Air Liquide is the clearest listed peer beneficiary through carrier gases and electronic materials. Its October 29 release independently reported 12% electronics growth and identified sustained AI-related semiconductor demand. Linde’s results therefore corroborate an industry-wide consumption signal rather than implying that Air Liquide will receive any portion of Linde’s $2 billion backlog. The earnings channel is higher supply volumes, utilization of existing gas assets and subsequent contributions from Air Liquide’s own contracted investments. For TSMC, the read-through is physical support for semiconductor manufacturing activity and the development of the U.S. supply ecosystem. Linde’s existing relationship and Phoenix infrastructure are directly relevant. However, Linde’s consolidated electronics growth cannot be treated as TSMC’s wafer-growth rate, and gas infrastructure does not establish that every fab has completed tool installation, qualification or customer production ramps. The non-consensus discipline concerns semiconductor capital equipment. Gas growth exceeding a broad wafer-area benchmark does not, by itself, establish stronger equipment orders or a quantified increase in process intensity. Geography, pricing, customer mix, existing-fab utilization and new projects can all contribute. Similarly, Linde’s selective approach to memory contracts is a capital-allocation preference, not evidence that memory investment is collapsing. TSMC had already reported on October 15, and Air Liquide reported on October 29. The immediate market significance is corroboration rather than a new forecast of those quarters. The longer-duration catalysts are electronics project commissioning during 2027-2029 and sustained gas consumption at customer facilities. Cross-portfolio exposure remains partly tied to the same semiconductor investment cycle, even when the securities sit in different sector classifications. HELIUM: HIGHER SELLING PRICES DO NOT ESTABLISH A DURABLE SCARCITY PROFIT STREAM (READ-THROUGH 5) Affected companies: Air Products and Chemicals (APD: United States) and Air Liquide (AI: France). Direction: negative for assumptions of persistent helium scarcity pricing and exceptional profit conversion. Magnitude: potentially meaningful within helium, but limited and unquantified at the consolidated group level. Call evidence: Linde stated that helium pricing increased across markets, but that the increases recovered sourcing and logistics dislocation costs. Management said it had not pursued excessive pricing and warned that “longer term, helium will be long.” Qatar’s potential recovery to approximately 75% of production was distinguished from the ability to transport that production to customers. The transmission mechanism operates through both delivered supply and cost. Longer alternative routes consume more container capacity and extend turnaround times. A producer can restart output without an equivalent increase in delivered helium if logistics remain constrained. During that interval, higher realized prices can largely reimburse extraordinary costs rather than expand margins. For competing helium suppliers, the high-conviction negative is to the durability of the scarcity-profit assumption, not a claim that their current earnings must fall. Source portfolios, procurement costs, inventories and customer contracts differ. Linde’s experience is insufficient to calculate a peer EPS adjustment or to conclude that every supplier faces the same price-cost spread. Normalization also has opposing effects. Lower logistics costs and improved availability can support volumes and cash conversion, while declining scarcity premiums or surcharges can reduce selling prices. The relevant forecast variable is retained gross profit after both movements, not the direction of helium prices alone. Long-term customer contracts support relationships and volume visibility but do not necessarily lock in today’s price. Air Products’ November 5 report is the nearest opportunity to distinguish helium volume, pricing and cost effects. For Air Liquide, subsequent operating updates can clarify whether improved supply benefits volumes more than normalization affects realization. The longer-duration issue extends into 2027 and beyond, but the call provides no reliable normalization date. Linde’s new customer commitments also should not be described as demonstrated share losses for either competitor. No displaced supplier was identified. The actionable conclusion is narrower: a permanent helium windfall is not a sound basis for raising long-term gas-sector earnings expectations. EUROPEAN CHEMICALS EUROPEAN INDUSTRIAL RECOVERY: FLAT PHYSICAL ACTIVITY AND COST DISADVANTAGES LIMIT THE REBOUND CASE (READ-THROUGH 6) Affected companies: LANXESS (LXS: Germany) and BASF (BAS: Germany). Direction: negative for European production economics and assumptions of a broad volume-led recovery. Magnitude: moderate to high for exposed European assets; more diluted at BASF’s diversified group level. Call evidence: Linde’s reported European volume growth rounded to 1%, but management attributed the increase largely to equipment sales. Excluding equipment, underlying activity was flat. Lamba stated: “There is, at this point in time, no catalyst.” He highlighted energy-cost volatility, continental Europe’s structural disadvantages and continuing chemical-industry restructuring and plant closures. The transmission mechanism is the combination of weak fixed-cost absorption and an unfavorable delivered-cost position. Flat demand limits operating leverage, while expensive or volatile energy reduces the competitiveness of European production against advantaged locations. Passing through input inflation can protect unit economics without restoring lost volumes or improving utilization. LANXESS is particularly relevant because its own Q2 disclosure identified high energy costs and inventory reduction as pressures on Advanced Intermediates. Its management also stated that there was no sustained upturn in core-market demand. Linde’s later-quarter observations reinforce that assessment and increase the burden on cost reductions rather than a cyclical recovery to deliver earnings improvement. BASF requires a more qualified interpretation. Its October 28 results showed higher earnings and a raised annual EBITDA outlook. Those results are important counterevidence against treating European weakness as a uniform deterioration across the group. Linde’s negative read-through applies most directly to European asset utilization, restructuring requirements and the sustainability of a volume-led regional recovery, not to a blanket downgrade of BASF’s consolidated earnings. The most actionable near-term event is LANXESS’s November 5 report, particularly the balance between price recovery, genuine volume improvement and cost savings. For BASF, Linde’s call mainly tests the durability and geographic composition of already-reported improvement. It does not negate that improvement. The longer-duration implication is continued capital allocation toward more competitive production locations and a higher restructuring burden for disadvantaged European capacity. Linde’s favorable distinction between Nordic and continental European activity also argues against treating Europe as a homogeneous industrial market. A broad European cyclical position requires evidence of demand recovery and competitiveness improvement that this call does not provide. U.S. CONSTRUCTION AND METALS DATA-CENTER CONSTRUCTION: PHYSICAL ACTIVITY SUPPORTS BACKLOG EXECUTION, NOT JUST PROJECT ANNOUNCEMENTS (READ-THROUGH 7) Affected companies: EMCOR Group (EME: United States) and Comfort Systems USA (FIX: United States). Direction: positive. Magnitude: material for the relevant construction businesses, but modest incremental information after their strong reported results. Call evidence: Linde reported low-teens U.S. packaged-gas growth associated with construction and data-center activity. Manufacturing sales increased 7%, and management separately identified stronger U.S. volumes. Packaged gases are used in fabrication, welding and installation activities. Their demand therefore provides a different observation point from hyperscaler spending announcements or contractor backlog awards. The signal supports the proposition that projects are generating physical work rather than remaining entirely in planning and procurement. EMCOR and Comfort Systems are exposed through electrical and mechanical installation, cooling infrastructure and related construction services. The transmission mechanism is execution of contracted work into recognized revenue, workforce utilization and eventual cash receipts. It is not an automatic increase in new orders: consumables used during construction can corroborate work already included in backlog. The distinction matters because both companies had already reported substantial growth. EMCOR’s October 29 release showed 25.5% organic revenue growth and $18.57 billion of remaining performance obligations, with the largest increase in network and communications projects requiring electrical infrastructure and sophisticated cooling. Comfort Systems’ October 22 release reported $17.81 billion of backlog and a constructive outlook into 2027. Linde adds independent confirmation of activity, not a new estimate of those disclosed backlogs. The gas growth rate must not be mapped directly onto construction volumes. Pricing contributed to packaged-gas sales, and the category includes activity beyond data centers. Nor does active construction prove that projects have received final utility connections or entered commercial operation. Near-term catalysts are Q4 execution, backlog conversion, labor productivity and cash collection. Over 2027, the fundamental opportunity is continued delivery of complex electrical and mechanical infrastructure. The cross-portfolio risk is that semiconductor gases, data-center contractors and AI hardware can share the same underlying investment cycle despite appearing in different equity sectors. This is a stronger signal for specialized infrastructure contractors than for a broad-based rebound in all commercial construction. U.S. STEEL: REGIONAL PHYSICAL DEMAND IS STRONGER THAN A GLOBAL CYCLICAL AVERAGE SUGGESTS (READ-THROUGH 8) Affected company: Nucor (NUE: United States). Direction: positive for the domestic demand and utilization backdrop. Magnitude: moderate operating relevance, with limited incremental earnings-surprise information. Call evidence: Linde’s metals and mining sales increased 3%, led by the Americas, with stronger North American construction and industrial activity. Management separately identified increased U.S. metals production, while describing slight contractions in EMEA and APAC and citing weaker Chinese steel production. The transmission mechanism is regional production activity. Industrial gases are consumed during steelmaking and related processing, so stronger U.S. demand provides an independent operating indicator for domestic producers. For Nucor, higher shipments and utilization can support fixed-cost absorption across mills, while construction activity supports downstream steel products. The read-through is corroborated by Nucor’s already-reported Q3 performance, which attributed stronger steel-mill and steel-product earnings to higher volumes and realized prices. It should therefore be treated as confirmation of regional strength rather than a new prediction of the completed quarter. The important limitation is that gas consumption is not a steel-margin measure. Steel prices, scrap and metallics costs, energy expense and product mix remain decisive. Linde’s own power-cost commentary highlights a potential offset for electricity-intensive steel production. Positive production signals do not justify translating a gas-volume increase into a proportional Nucor earnings increase. Near-term attention should remain on Q4 realized pricing and normal seasonal volume changes. Nucor already indicated that Q4 earnings could be broadly stable or modestly higher despite seasonally lower volumes, so Linde’s call does not support assuming sequential volume acceleration. The longer-duration implication is regional differentiation: U.S. manufacturing and infrastructure investment can sustain domestic demand even when European industrial activity and Chinese steel output are weak. That favors evaluating Nucor through its actual North American markets rather than applying an undifferentiated global materials-cycle forecast.

A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. Palantir $PLTR and Armada announced a partnership to deliver sovereign AI on modular data center infrastructure manufactured in the U.S. and allied nations. Under the agreement, Palantir named Armada its inaugural Certified Modular Data Center Partner, combining Palantir’s Sovereign AI Operating System with Armada’s Galleon modular data centers, Armada Platform, and Sovereign AI Grid. The joint offering allows enterprises and governments to deploy open-weight AI models on infrastructure they own and control, with the ability to fine-tune models on proprietary data while keeping compute, data, and hardware entirely within their own security boundary. The companies said the modular approach can bring new AI capacity online in months instead of years, addressing growing demand for sovereign AI infrastructure. 2. Nebius $NBIS has acquired Inferize, an AI inference optimization startup, for an undisclosed amount. Inferize’s technology is designed to reduce model “cold starts,” cutting the time GPUs remain idle while models load and new inference capacity comes online. Nebius plans to integrate the technology and team into its Token Factory inference platform to improve GPU utilization, lower inference costs, and serve more AI workloads using the same infrastructure. 3. NVIDIA $NVDA and SoftBank have each completed the final $10B installment of their respective $30B commitments to OpenAI’s latest funding round. The financing totaled $122B in commitments at an $852B valuation and included Amazon’s $50B investment. SoftBank’s total investment in OpenAI has now reached approximately $64.6B, giving it an estimated 13% ownership stake. According to The Information, OpenAI is already targeting another $30B funding round at a valuation of around $1.4T. 4. Netflix $NFLX co-CEO Ted Sarandos said the company “isn’t growing as fast as I want us to,” while noting that some of Netflix’s strategic decisions are intentionally creating headwinds to subscriber growth. He also revealed that live programming accounts for roughly 5% of Netflix’s $20B annual content budget but generates only about 1% of total viewing, underscoring the company’s continued focus on balancing long-term strategic investments with overall engagement. 5. Nike $NKE reported Q1 FY2027 results with EPS of $0.48, beating expectations of $0.43, while revenue missed at $11.21B versus $11.32B expected, down 4% YoY. Gross margin came in stronger than expected at 42.8%, helped by lower warehousing and logistics costs, but weakness persisted across NIKE Direct (-8% YoY), Digital (-13% YoY), and Greater China (-22% YoY). The biggest disappointment was guidance, with Nike forecasting high-single-digit revenue declines for FY2027 and adjusted EPS of $1.15–$1.35, well below the $1.65–$1.67 consensus. Management said it continues to see challenges in Sportswear, Jordan Brand, and Greater China, while rolling out its new Pace operating model to accelerate its long-term turnaround. 6. Broadcom $AVGO has agreed to provide Anthropic with up to $42B in financing tied to its AI infrastructure expansion, according to Anthropic’s IPO prospectus reviewed by Reuters. The facility could fund roughly one-third of Anthropic’s $125.2B five-year TPU compute commitment, and the debt may be convertible into Anthropic equity. Anthropic is also expected to become Broadcom’s largest compute customer in 2027. The filing notes that Broadcom’s dual role as both a hardware supplier and financing partner could create potential conflicts related to pricing and access to compute. 7. Amazon $AMZN plans to expand its in-house delivery network from roughly 23,700 U.S. ZIP codes today to about 39,400 by 2029, according to an internal planning document reviewed by Business Insider. That expansion would allow Amazon to reach approximately 95% of U.S. ZIP codes and 99.9% of customer demand. The plan includes building 165 new delivery stations between 2027 and 2029 and adding roughly 4,500 ZIP codes through its Hub Delivery rural network. The initiative reportedly began under Project Aurelian, a contingency plan created when negotiations with the U.S. Postal Service became uncertain, though Amazon said the projections remain preliminary and could change. 8. The top 10 most active options today by contracts traded were $NVDA with 2.6M contracts, $MU with 1.5M contracts, $TSLA with 1.3M contracts, $AAPL with 1.1M contracts, $SPCX with 786K contracts, $AMZN with 640K contracts, $GOOGL with 637K contracts, $INTC with 582K contracts, $NKE with 561K contracts, and $MSFT with 447K contracts. 9. Oracle $ORCL has reportedly signed a 5-year, $7B AI compute agreement with Tencent, according to the Financial Times. Under the deal, Oracle will provide access to roughly 100,000 advanced AI chips across its Southeast Asia data centers, with about 30% of the contract value paid upfront. Tencent plans to use the capacity to train AI models and expand its portfolio of agentic AI products, as the company accelerates AI infrastructure spending. The investment has already impacted Tencent’s financials, with Q2 infrastructure capex surging 176% YoY to $7.9B and free cash flow turning negative at -$2B for the first time in more than a decade. 10. AI-related companies continue to dominate capital markets, now accounting for roughly 40% of the S&P 500's total market capitalization. In emerging markets, just 3 semiconductor companies make up 28% of the MSCI Emerging Markets Index, while AI-related issuers have represented 49% of all investment-grade bond issuance so far in 2026. The concentration is even more pronounced in private markets, where 87% of year-to-date venture capital funding has gone to AI companies. For comparison, during the 1999 Dot-Com bubble, internet companies accounted for less than 40% of venture capital funding, highlighting the unprecedented concentration of investment flowing into AI today. 11. Since 1928, the S&P 500 has gained an average of 2.1% during the first half of October in U.S. midterm election years, making it the strongest-performing half-month of the fourth quarter. Positive returns occurred 71% of the time over that period, the third-highest hit rate of any half-month in Q4. The seasonal strength follows a historically weak second half of September, which has averaged a 1.1% decline. Momentum has also tended to remain positive later in the quarter, with the second half of October averaging a 0.3% gain and the first half of November delivering an average 1.7% return with a 79% positive hit rate. 12. Bond ETFs have attracted $23B in inflows over the last 5 trading days, accounting for 62% of all ETF net inflows despite representing only about 15% of total ETF assets. Fixed-income funds also made up 6 of the 8 most-purchased ETFs over the period, with investors allocating capital across cash, municipal bonds, high yield, Treasuries, and $TLT. The broad-based demand suggests investors are increasingly positioning around the bond market as yields remain near multi-year highs and volatility persists across asset classes. WALL STREET IS THE GREATEST SHOW ON EARTH.

Idle USDC earns while it waits. Supply from the agent wallet into curated lending vaults on BNB Chain and Base: Aave, Morpho, Lista, Zircuit. Gas is on Q402. Withdraw from the same wallet when the agent needs it back.