Cathie Wood's Insights on Meta's AI Potential

Meta stock is emerging as a compelling investment within the artificial intelligence (AI) sector.

Morgan Stanley has projected that four leading companies -- Microsoft, Amazon, Alphabet, and Meta Platforms (META) -- will allocate a staggering $300 billion toward data center infrastructure and chips by 2025, as they pursue their AI ambitions.

According to Cathie Wood, the founder of Ark Investment Management, the software sector could become the next frontier for AI investments. She anticipates that software companies may produce $8 in revenue for every $1 spent on chips from suppliers like Nvidia.

If Wood's predictions hold true, these tech giants stand to make substantial profits from their AI-focused spending, especially Meta Platforms. Here, we delve into why investing in Meta stock could be a smart move by 2025.

Meta's Llama 4: A Game-Changer in AI Models

Meta oversees popular social networks including Facebook, Instagram, WhatsApp, and Messenger, collectively engaging nearly 3.3 billion users worldwide each day. AI has become increasingly critical in enhancing user experiences across these platforms.

Meta has developed an AI-driven content recommendation system that learns user preferences, shaping their feeds on Facebook and Instagram. This innovation led to an 8% increase in user engagement on Facebook and a 6% rise for Instagram, translating into more ad impressions and higher revenue for the company.

Additionally, Meta is rolling out AI features like Meta AI, a virtual assistant that can generate text and images and participate in group chats. The quality of such an AI assistant depends heavily on the underlying large language model (LLM), and Meta has a rich data reservoir from billions of social media interactions to create an impressive suite of LLMs known as Llama.

While many leading LLMs (from creators like OpenAI and Anthropic) are proprietary, Llama is open-source, allowing a vast number of developers to access and improve its capabilities, aiding rapid debugging and feature enhancement.

The latest iteration of Meta's LLM, Llama 3.2, is set to be succeeded by Llama 4 in 2025, which Zuckerberg claims will be the industry's most advanced model. This marks a significant achievement for Meta, particularly considering the competitive landscape where companies like OpenAI initially held a lead.

Monetization Strategies at Meta

Meta reported a record revenue of $40.6 billion in its most recent quarter (ending September 30, 2024). A large portion of this revenue stems from advertising, showcasing the company's expertise in monetizing new features effectively.

Initially, Meta integrated advertisements into content feeds to capture users' attention while they browsed posts. Video formats like Stories and Reels have also significantly contributed to revenue, even though competitors like Snap and TikTok introduced similar formats earlier.

Meta AI has already attracted over 500 million monthly active users and, while currently free, the company sees potential for future monetization. For instance, businesses may pay to include links to their sites or products in responses from Meta AI.

Looking ahead, the synergy between Meta AI and Business AI could allow every business using Meta applications to create a tailored AI virtual agent capable of handling customer inquiries and even processing sales. Such a service could be monetized through subscription models or per-message fees, providing additional revenue avenues.

Meta's Favorable Valuation Amidst AI Competitors

Although Meta's stock surged 532% from a low in 2022, its trailing 12-month earnings per share (EPS) of $21.23 results in a price-to-earnings (P/E) ratio of just 27.2. This places it at a 15% discount to the Nasdaq 100, which stands at a P/E of 32.2, making it cheaper than other AI stocks valued over $1 trillion, except for Alphabet, which currently faces regulatory challenges.

Meta’s stock appears even more attractive based on forward estimates. Wall Street predicts an EPS of $25.33 for 2025, implying a forward P/E of 22.7. This means the stock would need to increase by 42% in the coming year just to align with the Nasdaq 100's P/E ratio.

Llama's development is crucial for Meta's future within the AI sector, as effective AI software relies heavily on quality LLMs. However, building advanced LLMs is capital-intensive. Training Llama 3 required approximately 16,000 of Nvidia’s H100 graphics processors, while Llama 4 may need over 100,000 GPUs.

In fact, Meta is set to invest up to $40 billion in AI infrastructure for 2024, with projections of $52 billion for 2025 according to Morgan Stanley. If Wood's projections come true, this substantial investment could yield hundreds of billions in revenue for Meta, promising remarkable returns for investors. Given its competitive valuation, Meta stock could emerge as one of the top choices among AI giants to consider investing in next year.

Meta, AI, Investment