KK Blog

Menu

Archives

2026 / 07

"AI 建议的信心陷阱：错误答案如何压缩我不知道的空间"

2026-07-20

解读 #AI建议#自动化偏差#批判性思维#决策#AI信心#人机协作#可验证性
"五项实验显示，错误 AI 建议会降低准确率并抬高信心。真正可执行的应对，是把独立判断和外部核验做进工作流。"
"The AI Advice Confidence Trap: Why Wrong Answers Suppress 'I Don't Know'"

2026-07-20

解读 #AI Advice#Automation Bias#Critical Thinking#Decision Making#AI Confidence#Human AI Collaboration#Verification
"A five-experiment preprint shows how wrong AI advice can reduce accuracy and inflate confidence. The practical answer is a three-stage verification l
"AI Agent 什么时候需要自己的 DSL：一套最小决策框架"

2026-07-20

最佳实践 #AI Agent#领域特定语言#LLM#Agent可靠性#DSL#可验证性#软件架构
"判断 AI Agent 何时需要领域特定语言，何时 JSON Schema 已经够用，并用最小 DSL 架构建立可验证的执行边界。"
"When Should an AI Agent Get Its Own DSL? A Minimal Decision Framework"

2026-07-20

最佳实践 #AI Agents#Domain-Specific Language#LLM#Agent Reliability#DSL#Verification#Software Architecture
"Learn when an AI agent needs a domain-specific language, when JSON Schema is enough, and how to build a minimal DSL with deterministic validation."
MCP Elicitation：把 Agent 的暂停、追问与恢复变成协议能力

2026-07-19

最佳实践 #MCP#MCP Elicitation#AI Agent#Human in the Loop#Agent安全#OAuth#Model Context Protocol
MCP Elicitation 把工具执行中的人工决策变成可恢复的协议状态。本文讲清 form、URL、安全边界、状态持久化与重试设计。
MCP Elicitation Explained: Form, URL, and Recoverable Human-in-the-Loop Workflows

2026-07-19

最佳实践 #MCP#MCP Elicitation#AI Agents#Human in the Loop#Agent Security#OAuth#Model Context Protocol
MCP Elicitation turns missing input into recoverable control flow. Learn form vs URL mode, state handling, security boundaries, and retries.
隐式工作区与外部记忆：J-space 研究如何重画 Agent 的状态边界

2026-07-19

洞察 #Agent记忆#J-space#Anthropic#Context Engineering#Agent可观测性#长期记忆
J-space 负责单次推理中的概念广播，可靠 Agent 仍需上下文、检查点、长期记忆和外部遥测共同维持状态。
教师 AI 的测量缺口：如何补上从备课效率到学习成果的反馈闭环

2026-07-19

洞察 #教师AI#学习成果#教育AI#ChatGPT for Teachers#Claude for Teachers#因果评估
教师 AI 需要把节省时间连接到课堂实践、学生行为和持久学习，用完整因果链支持学校采购与产品改进。
浏览器指纹失效之后：行为检测、签名身份与误伤成本

2026-07-19

调研 #Browser Agent#Bot Detection#Cloudflare Precursor#FP-Agent#Web Bot Auth#AI安全
真实 Chrome 让静态指纹失去判别力。可靠的 Browser Agent 治理需要行为检测、签名身份与细粒度授权三层协同。
"Latent Workspace vs External Memory: How J-Space Redraws AI Agent State Boundaries"

2026-07-19

洞察 #AI Agent Memory#J-space#Anthropic#Context Engineering#Agent Observability#Long-Term Memory
J-space handles concepts inside one inference. Reliable agents still need context, checkpoints, long-term memory, and external traces.
Previous Next