arXiv:2606. 15708v1 Announce Type: new Abstract: Welcome to the ninth edition of the AI Index report.
Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.
The paper "The AI-Enabled Scientific Frontier" analyzes 2,507 head‑to‑head comparisons of AI versus other scientific methods across 27 disciplines from 2000 to early 2025. It finds that AI often outperforms traditional statistics but at higher computational cost, while in about a quarter of cases AI is both more expensive and less effective. Compared to scientific computing, AI usually underperforms but at lower cost, though since 2020 its performance has improved, now surpassing computing in over half of the comparisons.
The paper "AI in Science: Early Insights" analyzes AI’s impact on scientific work using data from 15 million Gemini interactions, 2,600 specialized AI models, and a survey of 600 scientists. It finds widespread AI adoption, complementary use of large language models and specialized tools, significant productivity gains of about seven hours per week, and a shift in research bottlenecks toward hypothesis backlog and verification needs. The study suggests AI can boost scientific productivity but its full effect depends on addressing new downstream challenges.
arXiv:2606. 19899v1 Announce Type: cross Abstract: This paper addresses a rapidly emerging policy challenge: how to generate and interpret credible evidence about the biological capabilities and risks of AI scientists, or agentic AI systems capable of autonomously or collaboratively performing multi-step scientific tasks.
arXiv:2608. 05173v1 Announce Type: cross Abstract: As AI capabilities advance, AI systems will pose greater risks to national security and potentially humanity as a whole.
arXiv:2606. 12683v1 Announce Type: new Abstract: Over the last decade, building human-level artificial general intelligence has moved from far-fetched speculation to being a concrete next-decade target for many of the largest AI organisations.