Google and Johns Hopkins researchers introduce Pro2 Agent, which has ML research agents first build a 'Problem State' capturing suspected bottlenecks, candidate fixes and confidence levels, then run single-variable experiments to update beliefs before committing code changes. It reaches a 76.0% medal rate across all 75 MLE-Bench tasks within 48 hours, the highest reported, and medals on 57.6% of MLE-Bench Lite tasks with its first solution versus 28.8% for MARS.
Johns Hopkins University
约翰斯·霍普金斯大学
Leading US research university that published this grid equipment supply chain and sustainability analysis.
Curated Intelligence
Rohan Paul (@rohanpaul_ai)T2·Agents
78 pts
IT之家 · 智能时代T2·Infra
72 pts
A Johns Hopkins study in Cell Reports Sustainability finds US electricity demand will grow 21% by 2030, driven by AI data centers, EVs and electrification, triggering shortages of copper, steel and nickel as early as 2027. Transformer supply gaps approach 15% over the next four years and other key grid equipment gaps reach 30%, with data centers accounting for roughly two-thirds of the projected shortfall, constraining new plant and data center interconnection while pushing up power bills.