Platform giants hold context edge over agentic AI developers
Agentic AI faces a strategic disadvantage against platform giants like Apple and Google due to a lack of deep user context. This limits the ability of general agents to provide proactive, personalized services compared to vertical applications.
Agentic AI uses significantly more tokens than chat interfaces, driving higher compute costs. Mark Zuckerberg stated that Meta's Muse app has 10 million setups and 25 million daily users, with a quarter making e-commerce purchases averaging $200. However, he noted that scaling requires compute costs to fall by 10x. A key strategic challenge for Anthropic and OpenAI is the lack of user context compared to Apple, Google, Amazon, and Meta. These platforms have access to emails, messages, and browsing history, enabling proactive agent suggestions. Anthropic and OpenAI only see interactions within their own agents, limiting their ability to infer user intent or provide personalized proactive services. This creates a barrier to horizontal agent adoption, as vertical applications may capture value before general agents can compete. The lack of deep user context puts Anthropic and OpenAI at a disadvantage against platform giants in building proactive, high-value agentic experiences.