# Mate in the Loop

> How AI is actually being adopted across maritime, logistics, manufacturing and energy — a bilingual read reviewed by Kenny Jung, 34 years in the industry. Every insight is drafted by AI, checked against its source by Codex, and published only after Kenny Jung approves it ("AI-drafted · reviewed by Kenny"). Augmentation, not automation.

Bilingual: Korean pages live at the root (/), English pages under /en. Each insight links its original source; facts are quoted from the source in English, and the summary and take are our own words.

## Insights (approved signals, newest first; English headlines)

- [Anthropic is making automated action-checking the default in most Claude apps, replacing per-action human approval.](https://mateintheloop.ai/en/insights/aitimes-2026-10-23-anthropic-auto-mode-default): 2026-10-23 · Other industries · AI agent · AI security · stage: Production · source: AI타임스 (https://www.aitimes.com/news/articleView.html?idxno=216579) · Korean: https://mateintheloop.ai/insights/aitimes-2026-10-23-anthropic-auto-mode-default
- [A Seoul public agency built an in-house assistant from 26 years and 11,835 cases of construction supervision, guiding each stage and checking documents for errors.](https://mateintheloop.ai/en/insights/aitimes-kr-2026-10-23-seoul-facilities-construction-ai): 2026-10-23 · Other industries · AI / ML · LLM · stage: Pilot · source: 인공지능신문 (https://www.aitimes.kr/news/articleView.html?idxno=42417) · Korean: https://mateintheloop.ai/insights/aitimes-kr-2026-10-23-seoul-facilities-construction-ai
- [Siemens' Fuse EDA agent now chains seven of its own tools to run electrothermal analysis on 3D ICs over an 8–10 hour run, exploring hundreds of design points.](https://mateintheloop.ai/en/insights/engineering-com-2026-10-22-siemens-agentic-3dic-eda): 2026-10-22 · Manufacturing · AI agent · LLM · AI security · stage: Research · source: engineering.com (https://www.engineering.com/siemens-expands-agentic-ai-for-3d-ic-design-workflows/) · Korean: https://mateintheloop.ai/insights/engineering-com-2026-10-22-siemens-agentic-3dic-eda
- [HD Hyundai and Anduril ran a 3.5-hour autonomous-operations demo on a converted 368-ton US Navy test vessel off Long Beach, with the Navy, Coast Guard and ABS watching.](https://mateintheloop.ai/en/insights/shippingnewsnet-2026-10-21-hd-hyundai-anduril-usv-demo): 2026-10-21 · Maritime & ports · AI / ML · AI agent · stage: Pilot · source: 쉬핑뉴스넷 (http://www.shippingnewsnet.com/news/articleView.html?idxno=72575) · Korean: https://mateintheloop.ai/insights/shippingnewsnet-2026-10-21-hd-hyundai-anduril-usv-demo

## Hub

- [Insights archive and adoption map (EN)](https://mateintheloop.ai/en/insights): filter by industry, technology and adoption stage
- [인사이트 (KO)](https://mateintheloop.ai/insights)
- [RSS feed](https://mateintheloop.ai/rss.xml)
- [JSON Feed](https://mateintheloop.ai/feed.json)
- [Glossary KO ↔ EN](https://mateintheloop.ai/en/glossary)
- [AI security hub](https://mateintheloop.ai/en/insights?technology=ai-security)
- [llms.txt (Korean headlines)](https://mateintheloop.ai/llms.txt) · [llms.txt (English headlines)](https://mateintheloop.ai/en/llms.txt)

## About

- [About Kenny Jung](https://mateintheloop.ai/en/about): 34 years from chief officer on container ships to fleet planning, product and AI engineering
- [Privacy](https://mateintheloop.ai/en/privacy)

## Optional

- [Harness Design for Long-Running App Development — Anthropic Engineering Summary](https://mateintheloop.ai/en/blog/harness-design-long-running-apps): How Anthropic's GAN-inspired multi-agent architecture breaks quality barriers in long-running autonomous coding agents
- [Agentic Workflow Patterns Validated in ContainerGenie.ai](https://mateintheloop.ai/en/blog/containergenie-agentic-workflow): Practical design principles for applying LLM/RAG to maritime logistics operations
- [4 Critical Gaps in Building Just-In-Time Port-Call Services](https://mateintheloop.ai/en/blog/just-in-time-port-call-playbook): Common bottlenecks in real-time collaboration platforms between liners and terminals
- [Maritime Data for AI/ML: Operational Consistency Is Harder Than Data Collection](https://mateintheloop.ai/en/blog/maritime-data-for-ai-ml): Domain consistency issues and mitigation patterns in building maritime AI/ML datasets