Meta is attempting to revitalize its artificial intelligence strategy by integrating Alexandr Wang’s Muse Spark model, though the company continues to face significant skepticism from investors and developers.

Key Points

• Meta invested over $14 billion to recruit Alexandr Wang and his Scale AI engineering team to overhaul its AI development efforts.
• The company launched the proprietary Muse Spark model in April, signaling a shift away from its previous focus on open-source Llama models.
• Meta’s stock has declined 18% over the past year, making it the worst-performing megacap tech stock despite 33% revenue growth in the first quarter.
• Advertising currently accounts for 98% of Meta's total revenue, fueling investor pressure to successfully monetize new AI-first products.
• Meta faces a credibility gap with developers following the underwhelming performance of Llama 4 and over $80 billion in losses from its metaverse division.

Why it Matters

Meta’s pivot to proprietary AI models represents a high-stakes attempt to diversify its revenue streams beyond digital advertising. The company must now overcome significant developer distrust and investor fatigue to prove it can compete with industry leaders like OpenAI and Google.