The META settlement is less important as a cash event than as a liability-reset: it reduces the probability of a catastrophic headline, but it does not remove the underlying mechanism that keeps regulators, plaintiffs, and advertisers focused on engagement-maximization. The real medium-term risk is product friction — any changes that reduce time-spent among younger cohorts would hit Instagram monetization quality before they show up in reported revenue, so the stock can still de-rate even after the legal overhang narrows. The closest second-order loser is SNAP/PINS, where the same youth-safety playbook can be copied with lower tolerance for scrutiny.

NVDA is the higher-beta catalyst because its print can validate or puncture the entire AI capex chain in one shot. The market is likely underpricing the feedback loop into power, networking, and data-center infrastructure: if demand remains supply-constrained, the winners are not just semis but also grid capex and utilities with large load-connection pipelines. If the guide implies digestion, the first place that cracks is the crowded AI ecosystem rather than the core GPU franchise.

SO is a slow-burn beneficiary, but the upside is regulated and path-dependent. AI load growth only becomes equity-positive when it turns into approved rate base, transmission spend, and forward load commitments; until then, it is mostly narrative support. The contrarian view is that the consensus may be too eager to extrapolate a multi-year power shortage into near-term earnings upside — permitting, interconnection queues, and rate-case timing can easily push the pay-off out 6-18 months.