Tagged articles

RCT

3 articles · Page 1 of 1

Jul 21, 2026 · Artificial Intelligence

Feeling 20% Faster but Actually 19% Slower: Who Gains and Who Loses with AI Coding?

Three recent studies—an RCT with 16 senior developers, a METR follow‑up, and a large‑scale Cursor analysis—show that AI coding tools can slow experienced engineers by 19% while boosting PR merge rates by 39%, with speed gains depending on task type, code‑base familiarity, and how the AI is used.

AI codingCursorPR merge rate

0 likes · 10 min read

Feeling 20% Faster but Actually 19% Slower: Who Gains and Who Loses with AI Coding?

FunTester

Jul 15, 2025 · Artificial Intelligence

Do AI Code Generators Really Speed Up Development? RCT Shows 19% Slowdown

A randomized controlled trial with 16 veteran developers working on 246 real open-source issues found that using generative AI tools such as Cursor Pro with Claude actually increased average task time by 19 %, contrary to participants’ expectations of a 24 % efficiency gain, highlighting limitations of AI in complex, multi-module development contexts.

AI code generationOpen-source ProjectsRCT

0 likes · 9 min read

Do AI Code Generators Really Speed Up Development? RCT Shows 19% Slowdown

DataFunTalk

Jan 20, 2023 · Artificial Intelligence

Practice of Causal Inference Based on Representation Learning: RCT Standards, Joint Tree‑Neural Modeling, RCT‑ODB Fusion, and Feature Decomposition

This article presents a comprehensive industrial‑level guide to causal inference using representation learning, covering proper RCT experiment design, joint modeling of tree and neural networks, fusion of RCT with observational data, and advanced feature‑decomposition techniques to mitigate bias.

Feature DecompositionRCTpropensity score

0 likes · 22 min read

Practice of Causal Inference Based on Representation Learning: RCT Standards, Joint Tree‑Neural Modeling, RCT‑ODB Fusion, and Feature Decomposition