Meta Platforms (META) Morgan Stanley Technology, Media & Telecom Conference summary

AI research and infrastructure

• FAIR team has driven open AI research for a decade, contributing to open-source tools like PyTorch.

• GenAI team formed by consolidating language experts, focusing on building advanced AI infrastructure with $60–$65 billion investment this year.

• Llama series, especially Llama 3, released with open weights, has over 800 million downloads and is widely used by developers and businesses globally.

• Meta AI assistant now serves 700 million monthly users, integrated into major apps except Threads.

• Focus is on scaling, personalization, and enhancing voice and multimodal capabilities in Meta AI.

Open-source strategy and business impact

• Open-source heritage attracts top talent and fosters innovation, safety, and auditability.

• Llama's open model enables global developer contributions and optimizations, benefiting Meta's ecosystem.

• GenAI tools empower 4 million advertisers and hundreds of millions of businesses to create personalized content and automate customer interactions.

• AI-driven content personalization increases ad relevancy and conversion rates.

• Llama 4 aims to deliver advanced reasoning, agentic use cases, and native multimodal interaction.

Consumer and business use cases

• Meta AI is used for contextual search, news summarization, and ongoing personalized conversations.

• Voice and personalization are key to future AI assistant experiences.

• Monetization of Meta AI is not a current priority; focus remains on utility and user experience.

• AI tools already drive significant business messaging revenue, especially via WhatsApp in emerging markets.

• Multimodal models will enhance business messaging by enabling image and audio understanding.