AI Mental Health Frontier AI 心理健康前沿 — 心理健康 AI 需要按风险分层部署 SAMHSA 最新政策报告给出一个实用提醒：心理健康 AI 的治理应由任务风险、人工监督和证据决定，而不是由界面是否流畅决定。
AI Mental Health Frontier AI Mental Health Frontier — Mental-Health AI Needs Risk-Tiered Deployment A new SAMHSA policy paper makes the deployment lesson concrete: mental-health AI should be governed by task risk, human oversight, and evidence—not by how fluent the interface feels.
AI Investment Frontier AI Investment Frontier — Credit Models Need an Explicit Fairness Frontier A new semi-structured credit model shows how investment AI can expose the trade-off between predictive lift, interpretability, and group fairness.
AI Signals and Reality Checks 智能体浏览器真正需要的是状态，而不是速度 Cloudflare 的 Kitesurf 让浏览器执行变得更轻、更短暂。真正的工程瓶颈转向跨运行保存状态、权限与恢复路径。
AI Signals and Reality Checks Agent Browsers Are Really Disposable State Machines Cloudflare's Kitesurf makes browser execution cheaper and more disposable. The operational bottleneck moves to preserving state, permissions, and recovery across short-lived runs.
AI Mental Health Frontier AI 心理健康前沿 — 情绪支持必须先写清升级契约 当年轻人开始把聊天机器人当作情绪支持入口，真正的部署问题不是模型会不会安慰，而是谁在何时接手、如何负责。
AI Mental Health Frontier AI Mental Health Frontier — Emotional Support Needs an Escalation Contract Recent reporting on youth use of AI chatbots makes the deployment lesson plain: emotional support is not safe until escalation, supervision, and accountability are explicit system contracts.
AI Investment Frontier AI Investment Frontier — Financial Advice Needs a Lifecycle Test A new study of AI financial advice shifts the investment-AI question from fluent answers to lifecycle outcomes, suitability, and governance.
AI Signals and Reality Checks AI 搜索正在向开放网络收取一笔分发税 AI 答案可能降低点击价值，却提高被选为信源的价值。内容团队需要测量引用、品牌需求和有效行动，而不只是搜索流量。
AI Signals and Reality Checks AI Search Is Becoming a Distribution Tax on the Open Web AI answers may reduce the value of a click while increasing the value of being selected as a source. Publishers need to measure citation, not only traffic.
AI Mental Health Frontier AI Mental Health Frontier｜心理健康 AI 的安全缺口，已经是部署问题 最新临床与政策研究显示，心理健康 AI 已经接触到处于困境中的人，而评估、升级与责任机制仍不均衡。
AI Mental Health Frontier AI Mental Health Frontier — The Safeguard Gap Is Now a Deployment Problem Recent clinical and policy work shows that mental-health AI is already reaching people in distress, while evaluation, escalation, and accountability remain uneven.
AI Investment Frontier AI Investment Frontier — Investment AI Needs a Decision Trace A new benchmark argues that investment agents should be judged by point-in-time evidence, investor profile, reasoning, decisions, and delayed outcomes—not returns alone.
AI Signals and Reality Checks Stripe 收购 OpenRouter：模型路由正在变成财务控制层 Stripe 收购 OpenRouter 的意义不只是拿下一个 AI 网关，而是把模型选择、Token 支出和故障切换变成可管理的财务控制面。
AI Signals and Reality Checks Stripe's OpenRouter Deal Makes Model Routing a Financial Control Layer Stripe's OpenRouter acquisition is less about owning another AI gateway than about turning model choice, token spend, and fallback policy into a finance-grade control surface.
AI Mental Health Frontier AI 心理健康前沿 — 青少年应用不能只多一个生成式功能 一篇最新快速综述梳理生成式 AI 在青少年心理健康应用中的使用，也把真正的门槛指向证据、边界与升级机制。
AI Mental Health Frontier AI Mental Health Frontier — Youth Apps Need More Than a Generative Feature A new rapid review of generative AI in youth mental-health apps points builders toward evidence, boundaries, and escalation—not novelty.
AI Investment Frontier AI 投资前沿：风险偏好也应成为组合模型的输入 一篇新论文把机器学习的收益预期放进结构化组合模型，提示投资 AI 不应只预测资产，也要理解决策者的行为与约束。
AI Investment Frontier AI Investment Frontier — Risk Preferences Are a Portfolio-Model Input A new paper combines machine-learned return expectations with a structural portfolio model. The lesson for investment AI is to model behavior and constraints, not just forecasts.
AI Signals and Reality Checks Open Weights Are Cheap; Reliable Local AI Is Not The latest open-weight releases lower the cost of trying local AI. They do not lower the cost of proving that a model is safe, stable, and worth operating.
AI Mental Health Frontier AI 心理健康前沿——评估必须跟着干预一起设计 一项最新 ACM 研究把评估设计放回 AI 心理健康干预的核心：不能只看回答质量，还要追踪安全、人工支持与长期结果。
