Meta, Amazon and Alphabet in AI spotlight

Meta Platforms shares rose 2.7% in afternoon trading on April 25 after the company announced a multibillion-dollar, multiyear partnership with Amazon (AMZN) to scale its "agentic AI" capabilities. The stock closed the session up 2.4% at $675.03.

Under the agreement, Meta will deploy tens of millions of AWS Graviton CPU cores. These will power real-time reasoning and code-generation workloads for its next-generation AI agents, making Meta one of Amazon's largest silicon customers.

The move is framed as a shift toward more energy‑efficient computing infrastructure. It addresses investor concerns that rising AI-related capital expenditures could pressure Meta’s operating margins.

The Amazon deal complements Meta’s "Efficiency 2.0" initiative, which recently included a 10% workforce reduction to prioritize high‑growth AI projects. Analysts cited in the report say the partnership secures compute capacity to support expectations for roughly 30% year-over-year total revenue growth in Meta’s April 29 earnings report.

Recent gains also follow an earlier move when eMarketer projected Meta would overtake Google as the world’s top digital‑ad seller by net revenue this year, with ad sales expected to reach $243 billion.

Meta’s main AI-driven growth engine is its Advantage+ advertising suite. Advantage+ uses generative AI, powered by Meta’s proprietary Llama model, to create and test multiple ad variations across Facebook, Instagram and WhatsApp in real time.

The system automatically optimizes campaigns for conversions, reducing manual work for marketers and aiming to improve return on ad spend. Stronger ad performance can drive higher engagement and larger ad budgets on Meta’s platforms.

This creates a feedback loop in which better AI tools attract more advertising dollars, which in turn fund further AI development. The article describes this as supporting Meta’s capital‑light, high‑margin business model.

Meta’s market value stands at about $1.7 trillion, with a recent share price around $675.49 and a 52‑week range between $520.26 and $796.25. The company also pays a dividend, with a yield of 0.31% cited in the data snapshot.

Bill Ackman’s Pershing Square Capital Management has roughly 38% of its $15 billion stock portfolio invested in three AI-focused companies: Alphabet, Amazon and Meta Platforms, according to the reports.

Ackman is described as a value investor who prefers concentrated, long‑term positions in businesses he views as mispriced, with durable competitive advantages and strong cash flows.

The three holdings are characterized as members of the "Magnificent Seven" that not only benefit from AI trends but help shape them. The articles highlight their combination of market leadership and what is described as still‑reasonable valuations.

Alphabet (GOOGL) has evolved from a search and advertising company into a vertically integrated AI provider. A core element is its Tensor Processing Units, custom chips built with Broadcom (AVGO) to train and deploy its Gemini AI models.

These TPUs power AI features across Google Search, Android and Workspace, and also support Waymo, Alphabet’s autonomous driving unit. The strategy combines control over silicon, models, data and distribution channels.

Google Cloud Platform is described as a high‑growth engine as enterprises move AI workloads to Alphabet’s infrastructure. Alphabet’s advertising businesses, including Search and YouTube, are also using machine learning to improve targeting and engagement.

Alphabet trades at a forward price‑to‑earnings multiple of about 29. The report notes that free cash flow is expected to compress this year as the company spends nearly $200 billion on AI‑related capital expenditures.

Amazon’s AI strategy centers on its Trainium and Inferentia chips, which support Amazon Web Services. The articles state these chips give AWS a structural cost advantage over rivals reliant on third‑party hardware.

AWS is positioned as a leading platform for hosting AI workloads, allowing enterprises to access and deploy models without building their own infrastructure. This underpins Amazon’s broader cloud growth.

AI also supports Amazon’s logistics and retail operations through robotics in fulfillment centers, predictive analytics for last‑mile delivery and computer vision to reduce shrinkage in physical stores.

On the consumer side, machine learning powers personalized recommendations and dynamic pricing, while Prime Video uses AI to refine content curation and targeted advertising. The articles emphasize that these initiatives reinforce one another across Amazon’s ecosystem.