Algorithmic efficiency—how quickly and resource-effectively an algorithm solves a problem—is foundational to computing but lacks a centralized way to track its progress over time. Currently, researchers and engineers rely on fragmented benchmarks or individual studies, making it hard to spot broader trends or predict future improvements. This gap makes decisions about algorithm choices, hardware design, and research priorities harder.

The Core Idea: Efficiency Trends Decoded

One approach is to create a data-driven map of how algorithms have improved across key fields like machine learning, cryptography, and compression over the past decade. This could take the form of a research paper analyzing published improvements or an interactive tool to explore trends visually. The project would involve gathering data from papers and benchmarks, standardizing metrics (e.g., accounting for hardware advancements), spotting patterns like diminishing returns, and possibly projecting future gains. For example, it might reveal that machine learning training algorithms improved exponentially until 2020 but are now plateauing.

Who Benefits and Why It Matters

Such a resource could help:

• Researchers identify overlooked areas for optimization.
• Engineers choose the most efficient algorithms for their needs.
• Hardware designers align chip architectures with algorithmic trends.
• Investors guide funding based on projected efficiency gains.

Incentives for participation could include academic citations for researchers or showcasing breakthroughs for tech companies. Open-source communities might use it to prioritize optimization efforts.

Getting It Off the Ground

Starting small could make this manageable. A minimal version might focus on one or two domains, like machine learning and sorting algorithms, using free data from sources like arXiv or MLPerf. Later phases could expand to more domains, build interactive features, or add predictive modeling. To address challenges like inconsistent data or hardware’s influence, early tests could compare algorithms within one domain while adjusting for hardware improvements. Over time, the project could offer custom reports or API access as potential revenue streams while maintaining transparency—for example, by highlighting industry contributions without letting them bias the data.

By systematically tracking algorithmic progress, this idea could reveal insights often overshadowed by hardware advancements, helping the tech community make smarter, data-backed decisions.