GPT-4 forecast the results of 70 real survey experiments almost as well as running them — and a new Science agent now runs biomedical studies on its own. What autonomous research instruments mean for your lab.

New AI systems from Google DeepMind and FutureHouse generate hypotheses, design experiments and interpret results — but the evidence still comes from the lab bench. Plus deepfake face detection, deep learning for company dynamics, and the evolvable-AI threat debate.

A four-study look at when humans take AI financial advice, a blueprint for governing AI in peer review, and new tools for doing inference with machine-generated variables.

A new ISR study shows why feeding AI-generated variables straight into your regressions biases the results — plus a deep-learning ECG biomarker, agentic GenAI in operations, and meta-learning for consumer preferences.

Early evidence that AI tools may erode the skills they were meant to augment, the risk of a scientific 'monoculture,' an autonomous medical AI agent, and new work on AI-assisted creativity and algorithmic transparency.

Early evidence that AI may be dulling the skills it was meant to sharpen — set against a 791-person field experiment at P&G where AI worked like a teammate, plus a paradox in how AI democratizes knowledge.

A 41-million-paper study on AI's double-edged effect on science, an autonomous medical AI that out-diagnoses physicians, and what AI agents mean for research.