2026-08-18
AI 快讯日报.

Warp 推出 Warp Factories：面向 AI 开发的即装即用软件工厂系统 ↗

On Tuesday, Warp introduced Warp Factories, a new infrastructure system designed to make building AI software factories as easy as possible.

OpenAI 推出面向青少年的 ChatGPT，内置学习模式与家长控制 ↗

ChatGPT for Teens adds age-appropriate safety measures, parental controls, and learning tools designed to steer teens away from harmful content — and from using AI to cheat on their homework.

Google 开源 SAM：为 AI 智能体打造的零配置、零信任 P2P 网络 ↗

google/sam is not Segment Anything. SAM here means Sovereign Agent Mesh, an Apache-2.0 networking project for autonomous AI agents. The problem it targets is concrete. Agents now run across cloud servers, on-prem datacenters, laptops, Raspberry Pis and Android devices. Letting them share tools usually means exposing internal scripts, LLM endpoints or private APIs to the public internet. SAM’s alternative is a zero-config, zero-trust P2P overlay — closer to a private VPN, but scoped to agent-to-agent tool sharing over the Model Context Protocol.

Anthropic CEO 称 AI 天然趋向集中，开源模型只是把权力转移给芯片持有者 ↗

An open fight over AI regulation has broken out on X. Investor Gavin Baker, former White House adviser David Sacks, and Meta researcher Yann LeCun accuse Anthropic CEO Dario Amodei of using fear rhetoric to buy himself a regulatory advantage. Amodei counters that regulation can also rein in corporate power, and that open models alone just shift power toward the players with the most computing muscle.

Firefox Smart Window 更新：AI 浏览模式接入实时网页信息与来源链接 ↗

Firefox’s Smart Window mode uses AI to organize selected tabs and find the important info from your browsing history to keep you on track.

谷歌以1000万美元拍下破产航司Spirit数据，用于AI训练 ↗

Google Pet Memory 认不出我的猫：Gemini for Home 宠物识别功能实测失败 ↗

Gemini for Home promises its Nest cameras can recognize my cats by name — but its AI can’t tell them apart, making the notifications useless.

Alvys 推出面向货运 TMS 工作流的 AI 智能体平台 Alvys Foundry ↗

Checking the site connection security

JAMA 评论文章：AI 超越医生时，监管者不应强制人类介入 ↗

An opinion piece in the medical journal JAMA argues that autonomous AI will soon outperform any doctor-AI team at medical reasoning tasks. The authors warn against writing a doctor's final say into regulation, but concede that almost all the evidence comes from simulations, not real patient care.

美国司法部调查 Andreessen Horowitz 合伙人兼任竞争AI公司董事 ↗

Andreessen Horowitz is the focus of an antitrust probe by the US Justice Department. The charge is that the firm's partners sit on the boards of competing data firms Databricks and Fivetran at the same time. The venture capital firm's close ties to the Trump administration add a twist, since it has actively lobbied for the administration's AI deregulation.

OpenAI 推出面向青少年的 ChatGPT 版本 ↗

OpenAI is shipping a version of ChatGPT tailored to users aged 13 to 17.

ChatGPT 推出青少年专用模式 ↗

How is this not called ChatGPTeen?

OpenAI 与 CodeAI 合作推出 ChatGPT 青少年版，内置学习模式与家长控制 ↗

该产品专为 13-17 岁用户设计，内置学习模式和负责任作业提醒，同时提供家长管控功能，双方未来一年将推进五大合作项目普及 AI 教育。#ChatGPT for Teens#

OpenAI 推出 ChatGPT for Teens：面向青少年的学习体验与更强安全保护 ↗

We’re introducing ChatGPT for Teens, an experience designed to help teens learn, think critically, deepen understanding, and use AI with confidence. It provides stronger built-in safety protections for teens, including features to promote healthy use and additional controls for parents. If our system estimates someone is under 18 or they state their age is between 13 and 17,⁠ they are automatically placed into ChatGPT for Teens.

OpenAI 与 CodeAI 达成合作，为第一代 AI 学生提供教育资源 ↗

Today’s students will be the first generation to grow up with AI as part of everyday life. For parents and educators, the question isn’t simply whether young people will use AI, but whether they will learn to critically evaluate its outputs, understand its limitations, and use it responsibly.

沃尔玛零售员工的新职责：帮 AI 智能体"纠错" ↗

为了推动业务增长，沃尔玛正大举投入 AI，并向数千名前线员工提供新的工具和培训。采用任何新流程都需要一个学习过程，沃尔玛这一轮 AI 推广还有一个特殊之处：AI 智能体本身也需要接受训练，才能提高准确性。

Anthropic 在 Vercel AI Gateway 的 token 成本是其他厂商平均的 4.4 倍 ↗

Anthropic dominated Vercel's AI Gateway spending in July, pulling in 65.1 percent of total revenue while accounting for only 30 percent of tokens processed. Its tokens cost 4.4 times as much as those from competing providers on average.

Claude Code 新增 /design 命令，可在终端直接创建 UI 设计稿 ↗

With the /design command, Anthropic brings a visual design workflow directly into Claude Code. Developers can generate UI mockups as artboards right in the terminal before writing any code. Claude reads the existing codebase and matches the current UI style.

解读智谱 GLM-5.3 成绩单：头条数字之外的信号 ↗

Zhipu’srelease notefor GLM-5.3 contains a sentence that did not make it into most of the coverage. Describing its own cybersecurity results, the Beijing company writes that capability “is growing fastest exactly where we are furthest behind.”

基准测试末日：LLM 让刷分变得轻而易举 ↗

There's been a lot of talk about the vulnpocalypse, to which I don't have much to add because I'm not a security person, but I haven't seen much discussion on the closely related (and to be fair, less serious, issue), the benchmarkpocalypse.