Apple and Microsoft are steering toward opposite AI futures, with a fundamental dispute over whether users want autonomous productivity agents or contextual personal assistance. Microsoft’s "Project Solera" positions hardware as thin portals for enterprise-grade cloud agents, while Apple leverages its massive installed base to keep the iPhone central by prioritizing personal context and privacy-preserving retrieval over pure frontier model performance. The central tension pits Microsoft's reliance on enterprise-driven agentic work against the Apple-advocated view that consumer demand is overwhelmingly tilted toward convenience rather than productivity.

The conflict here is less about technology—both firms have the infrastructure to match each other—and more about who better understands their own customer base. Apple’s case is stronger because it monetizes an existing, high-trust relationship with consumers, whereas Microsoft’s model faces an uphill battle to turn "agent-first" hardware into a mass-market reality. You can skip the video; the summary covers all technical claims and the strategic divide without the keynote's marketing padding.