Open and closed models are on different exponentials

This article analyzes the diverging performance trajectories between open-source and closed-source AI models, suggesting they follow different exponential growth curves rather than converging paths. The analysis likely examines factors such as compute investment, data availability, and development resources that cause these model categories to scale at different rates. It explores implications for the future competitive landscape between open and closed AI systems.

Source: Interconnects | 2026-06-01