Open Dataset Measures AI's Actual Impact on Accelerating Scientific Discovery
soumitrashukla9 · x · 2026-08-12
User @testingham shares an open-source GitHub repository, ai-discovery-data, designed to track rates of discovery over time across various domains to verify if AI has actually accelerated scientific progress.
The dataset includes consistent, public, and rebuildable data from fields like vulnerability disclosures, mathematics, and algorithms. For instance, metrics show an accelerating trend in curl and Firefox vulnerability disclosures, while areas like data compression, mixed-integer programming (MIP), and the Stockfish chess engine show sparse data or no obvious trend changes.
More from Research
- Cornell Paper: Encrypting Chain-of-Thought Fails to Prevent Model Distillation — burkov · 2026-08-12
- Raspberry Pi Runs Robot Path Planning Algorithm, Covering Campus in 19s — 4310sy · 2026-08-12
- NBER Paper: GenAI Faces Oligopoly, Open Ecosystems Need a 'Rogue' Giant — soumitrashukla9 · 2026-08-12
- SiPE: Integrating Syntax Priors into Transformer Positional Embeddings — UArizona · 2026-08-12
- Study Finds: Higher Reasoning Effort Makes LLMs Less Likely to Ask Clarifying Questions — xuanalogue · 2026-08-12
- Six Months of AI Auto-Research Tools, But No Clear Acceleration in Algorithmic Efficiency — soumitrashukla9 · 2026-08-12