Tech stocks today: Amazon to acquire satellite company Globalstar as it takes on Elon Musk's Starlink

Amazon agreed to acquire Globalstar for $11.57 billion, a major satellite-network deal aimed at challenging SpaceX’s Starlink. AI remains the dominant theme: TSMC said Q1 revenue rose 35% on continued AI demand, Anthropic disclosed a $30 billion run rate, and OpenAI publicly criticized Anthropic’s growth accounting while expanding competition in cloud and chips. Broader tech sentiment was supported by easing Iran-war fears, though the article also highlighted elevated geopolitical and AI-related risk.

Analysis

The strategic center of gravity in AI is shifting from model quality to control points: compute, cloud distribution, and enterprise trust. The most important second-order effect is that the winners are increasingly the infrastructure vendors that can sell both scarcity and optionality — custom silicon, cloud capacity, and networking — while standalone model providers face margin pressure as they scale. That favors names tied to supply elasticity more than names tied to headline model momentum.

The competitive noise around revenue accounting is less important than the signal it sends: enterprise AI buyers are still in vendor-selection mode, which means procurement cycles will likely elongate before they normalize. That creates a near-term deceleration risk for the “pure software” basket, while actually improving pricing power for incumbents with distribution leverage and embedded enterprise workflows. A trust-and-governance pitch becomes more valuable in regulated verticals, but it is also easier to copy than compute capacity.

Satellite and power are the underappreciated beneficiaries of the same theme. If large platforms push into direct infrastructure ownership, the bottlenecks move into launch cadence, energy availability, and capex intensity, which can become a drag on ROIC before revenue scales. That argues for watching not just the obvious hardware winners, but also the enablers exposed to long-duration contracts and capacity monetization — especially if AI capex remains sticky over the next 12–18 months.

Contrarian take: the market may be underpricing how quickly hyperscaler-led infrastructure competition compresses returns across the AI stack. The bullish consensus assumes more demand simply lifts all boats, but if each major platform starts internalizing chips, clouds, and network access, third-party providers face a classic scale-vs-margin tradeoff. The upside is still intact, but the cleaner trade is to own bottleneck owners and selectively fade the most crowded software narratives on rallies.

Key Decisions for Investors

• Long TSM / short a basket of high-multiple AI software proxies over the next 4-8 weeks: TSM benefits from durable capex visibility while software names are exposed to procurement scrutiny and accounting skepticism. Favor a 1:1 or 1:1.5 risk-defined pair.
• Add to NVDA on any 3-5% pullback, with a 3-6 month horizon: the market is still underestimating the value of being the default integration layer even as customers explore custom chips. Use options if you want convexity, but size for volatility.
• Initiate a tactical long CRWV versus a short in lower-quality cloud/infrastructure names for 1-3 months: multi-year capacity contracts are exactly where pricing discipline should hold if AI demand stays tight.
• Avoid chasing BE after the power-agreement pop; prefer selling covered calls or trimming into strength. The more important trade is the second-order beneficiary set around power, not the headline-name rerating.
• Speculative long AMZN versus GOOGL over 3-6 months: AMZN has a clearer path to monetizing AI through infrastructure control and distribution, while GOOGL is more exposed to price competition in consumer monetization and ad friction.