Serenity, the small-cap stock-picker whose reconstructed views this site tracks, spent Monday reacting to two developments: a new AI product from Meta and a sharp rebound in crypto prices. Neither prompted a fresh trade, but both fed directly into his running thesis about where AI demand lands upstream.

Meta's "Muse" as a demand telegraph. Serenity said he came away impressed after testing Meta's newly released Muse tool, describing it as capable of handling long-form, multi-step tasks. His interest, characteristically, was less in the product than in the hardware it implies: agentic workloads that spin up many browsers and virtual machines pull on general-purpose compute and memory. He tied that to CPUs from Intel, AMD and Arm, plus the networking cards, SSDs and DRAM those workloads consume. Meta (META) sits in his framework as a profitable hyperscaler whose heavy AI spending, he argues, is generating real returns rather than vanishing — and one that helps fund his upstream semiconductor and photonics bets.

The CPU angle is central to his thinking. Serenity has long held Intel (INTC) as a policy-driven bet that Washington will not let its domestic foundry champion fail, and he pointed to recent comments from Intel's CEO that CPU demand is so strong the company can only meet roughly half of customer needs — a data point he reads as confirming the shortage. Arm (ARM), the UK chip-design firm whose architecture underpins hyperscaler CPUs like AWS Graviton and Google Axion, anchors his "inference era" view that CPUs reclaim share as AI shifts from training toward serving. AMD rounds out the CPU trio; he treats it mainly as a demand engine that pulls through his actual small-cap picks. He also wondered aloud whether Meta hosts Muse entirely in-house or leans on outside capacity from Amazon (AMZN) or Nebius (NBIS) — the latter his favored "Neocloud," an AI cloud-capacity provider he holds at a large weight and views as a future AWS-scale operator.

Crypto bounce as a risk-appetite tell. Separately, Serenity noted that Bitcoin and Ethereum staged unusually strong recoveries, and mused that crypto-linked names like Robinhood (HOOD) and Coinbase (COIN) may be finding a floor. He credited on-chain equities trading and regulatory exemptions as fuel for the sector. Robinhood is a name he has turned mixed on — bullish on its retail cross-selling, but critical of decelerating revenue and its failure to add international equities. Coinbase, the largest US crypto exchange, he values more for its institutional ETF infrastructure and stablecoin revenue than its trading business.

The broader point he drew was a market read rather than a crypto call: he treats digital assets as a leading indicator for high-beta AI equities, arguing they tend to front-run rallies and drawdowns. On that logic, he suggested overall risk appetite may be returning despite the rate-hike backdrop — a tentative, macro-level observation rather than a positioning change.

Neither post carried a disclosed trade. Both amounted to Serenity restating, through the day's headlines, his core habit of reading hyperscaler and market behavior as a signal for the upstream compute, memory and components his portfolio ultimately depends on.