2026-08-19 科技日报

扫描 566 篇候选内容 · 覆盖 650 个信息源 · 纳入 555 篇

📌 今日要点

• 34% of the US public is now aware of AI xrisk, and the curve is steepening
• Stripe declares we’re living in the singularity and uses it as a reason not to IPO
• Attackers are using AI to build exploits for industrial control systems, U.S. agencies warn
• OpenAI fixes Codex bug that deleted real user files without permission
• China lets Nvidia’s H200 chips trickle onto the mainland to help its AI firms keep pace with the US
• GLM-5.3 tops the open-model rankings and undercuts rivals on price, but its release is delayed

📝 科技简讯

• 34% of the US public is now aware of AI xrisk, and the curve is steepening — LessWrong
• Stripe declares we’re living in the singularity and uses it as a reason not to IPO — The Decoder
• Attackers are using AI to build exploits for industrial control systems, U.S. agencies warn — The Decoder
• OpenAI fixes Codex bug that deleted real user files without permission — The Decoder
• China lets Nvidia’s H200 chips trickle onto the mainland to help its AI firms keep pace with the US — The Decoder
• GLM-5.3 tops the open-model rankings and undercuts rivals on price, but its release is delayed — The Decoder
• AI labs are failing to keep their own systems in check — The Decoder
• Anthropic says any lab can now let a language model agent run the whole protein design stack — The Decoder
• VLCP: Vision Language Control Policy Closed-Loop Code Replanning for Robot Manipulation — ArXiv RO (cs.RO)
• FetchMan: Learning Visual Humanoid Loco-Manipulation Policies from Simulated Experiences — ArXiv RO (cs.RO)
• Lambda-Hold Control: Human-Like Movement Emerges from a Minimal Task Reward in Predictive Musculoskeletal Simulation — ArXiv RO (cs.RO)
• Terrain-Aware Local Path Planning with Global DEM Data Integration for Autonomous UGV Navigation — ArXiv RO (cs.RO)
• PDDL-ART: Autonomous Symbolic Abstraction From Demonstration For Long-Horizon Robotic Manipulation Using Vision-Language Models — ArXiv RO (cs.RO)
• Multi-Observer Vehicle Localization Case Study with Roadside Radar and Connected Vehicle Sensing — ArXiv CV (cs.CV)
• AerialYield-B2D: A Greenhouse Blueberry Dataset with Five-Stage Ripeness Masks and Fruit Counts — ArXiv CV (cs.CV)
• PXDepth: Pixel-Space Modeling for Structure Preserving Monocular Depth Estimation — ArXiv CV (cs.CV)
• YILDIZ-VPR: A Novel Dataset with Dense Coverage Under Diverse Environmental Conditions for Visual Place Recognition — ArXiv CV (cs.CV)
• The 10th AI City Challenge — ArXiv CV (cs.CV)
• Tiny robots powered by light can hunt down and collect bacteria — Science Daily AI
• VentureBeat names Rob Strechay as its first Lead Analyst, expanding its enterprise AI research push — VentureBeat AI
• Why Companies Are Turning to Salesforce as a Customer Success Platform — Salesforce AI Blog
• The Salesforce Architect’s Guide to Dreamforce 2026 — Salesforce AI Blog
• Is It Worth Upgrading to Agentforce for Financial Services? — Salesforce AI Blog
• You Already Own These AI Agents, but Are You Using Them? — Salesforce AI Blog
• An IT Leader’s Guide to Dreamforce 2026 — Salesforce AI Blog
• LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation — Hugging Face Blog
• 5 new ways to level up your learning with Search — Google AI Blog
• How to Answer AI System Design Interview Questions — KDnuggets
• 5 Tools for Building and Deploying AI Agents in Production — KDnuggets
• Proactive Road Safety Intervention in Australia: Predicting Risky Driving Hotspots from Connected Vehicle Data — ArXiv ML (cs.LG)
• Detecting and Discriminating Operator Misspecification in Hybrid PDE-Parameter Learning: a Reference-Free Instrument, with Discrimination Bounded In Sample — ArXiv ML (cs.LG)
• Data-DPO: Direct Preference Optimization for Target Model Data Selection in LLM Post-Training — ArXiv ML (cs.LG)
• Hierarchical Data Selection via Manifold Coverage and Sparse Feature Coverage in LLM Post-training — ArXiv ML (cs.LG)
• Benchmarking Classical and Transformer-Based Models for Document Sensitivity Classification — ArXiv ML (cs.LG)
• Mr.Dec: Daily-Scale Longitudinal Multimodal Modeling for 30-Day Readmission Prediction — ArXiv ML (cs.LG)
• EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning — ArXiv ML (cs.LG)
• Domain and publish date filters for Web Search on AgentCore — AWS Machine Learning
• Automate Document Processing with Quick Automate and the IDP Accelerator — AWS Machine Learning
• Asynchronous patterns for calling Amazon Bedrock AgentCore agents in serverless pipelines — AWS Machine Learning
• How Fanatics Betting and Gaming built a multi-agent customer support system — AWS Machine Learning

📊 数据概览