AI Talent Market & Competition - The AI talent pool is very small, with perhaps a maximum of 2,000 people globally capable of building frontier models [2][7] - Competition for top AI talent among companies like OpenAI, Meta, Anthropic is intense, leading to potentially nine-figure or even eight-figure compensation packages [1][3] - Meta structures AI talent offers like executive pay, highly correlated to company and individual performance targets, not just typical restricted stock units [9][10] Company Performance & Strategy - Google is considered wildly ahead in AI among big tech companies, with a strong talent density and Gemini performing well [13][15] - Apple is significantly behind in AI talent and may need to acquire a company like Anthropic, which could cost over $50 billion [14][16][17] - OpenAI acquired Cursor, indicating the plug-and-play nature of AI models and the risk of disintermediation for companies only at the model layer [19] Future Trends & Challenges - Data, scale, and network effects are key to long-term lock-in in the AI space, similar to social media [22] - Persistent memory and data integration across devices and applications are crucial for AI models to achieve true lock-in effects [20][21]