Mapping Cost Trajectories of AI-Adjacent Technologies

The question

How quickly are the technologies that support AI, or will be transformed by it, improving? Historical costs and performance can help identify constraints on AI development and understand how capabilities might spread through the economy.

The project’s approach

The programme combines historical prices with technical specifications to construct technology cost and performance trajectories. Its methodology includes defining product categories, collecting and checking archival data, choosing performance measures, and estimating trends.

The intended datasets contribute to the Technology Observatory, an effort to build an empirical foundation for understanding technological change.

My participation

I’m participating in the Fall 2026 SPAR cohort under Isaak Mengesha’s mentorship. The work connects my interests in data, economic reasoning, and AI’s wider implications.

Related work

My completed Safe AI Germany project explored how uncertainty about AI’s future influences economic and career decisions.