HKA Chinese AI system called Qiushi Engine, developed by a team at Zhejiang University, has surpassed Anthropic’s Claude Code in an international benchmark for autonomous scientific research. As of Tuesday, Qiushi Engine ranked first on the ResearchClawBench leaderboard, followed by Open Science Desktop and Claude Code. The benchmark evaluates AI agents' ability to independently conduct research and compare their findings to human-authored papers. The test was created by researchers at the Shanghai Artificial Intelligence Laboratory to determine whether AI systems can truly perform the complex tasks they claim to handle. Qiushi Engine, launched recently, is described as a large language model-based agent capable of end-to-end autonomous scientific discovery in real-world settings.
Bias read (Center): The article reports on advancements in AI technology without taking a stance on political issues. It focuses on technical achievements and does not involve political figures, policies, or contentious topics.
