// Infrastructure

Source: Search Engine Journal

The proliferation of competing technical standards—Model Context Protocol, Agent-to-Agent communication, Natural Language Web, and Agents.md—reflects an infrastructure moment where no single vendor has locked in dominance over how AI systems will interoperate and delegate tasks. Unlike previous platform wars fought over closed ecosystems, these standards battles are being conducted in the open because economic value accrues to whoever controls the interoperability layer itself, not the endpoints. Whichever protocol stack wins determines whether AI agents become modular, composable tools that shift power to end-users, or proprietary black boxes that concentrate control among a handful of model providers.

Source: The Wall Street Journal

Both companies are projecting profitability to investors while obscuring a structural problem: computational costs to run their models post-training now exceed 50% of revenue, compressing margins to unsustainable levels at scale. This explains the simultaneous push for cheaper inference optimization, longer context windows to reduce repeat queries, and cache-heavy architectures—not as product features, but as operational necessity. The gap between board presentations and the physics of their cost structure suggests either a dramatic breakthrough in inference efficiency or a reset in pricing expectations within 18-24 months.

Source: Exponential View