AI
A conversation on Public Diplomacy Examined (IPDGC, GW) about what AI changes in diplomacy — and what it makes more valuable than ever.
AI
I just finished Dan Spokojny’s great article “Should Diplomats Trust AI?” The piece is anchored by two new items: the recent Foreign Service Journal issue that focused heavily on AI (which I contributed to), as well as the December 2025 AI-Powered Diplomacy paper from the Pavel Slunkin writing
The skills that got you promoted are often the ones that will undermine you next. This is a guide to the three career transitions that change what effectiveness means — from individual contributor to manager, from manager to leader of leaders, and the inner work that makes the difference.
What This Is: A real-world guide to building a personal knowledge management system that actually works with AI and turns Claude from a chatbot into a genuine thought partner. The Real Problem I'm constantly synthesizing research, connecting patterns across projects, and developing frameworks that integrate insights from
AI continues to grow at a dizzying pace. This beginner's guide helps diplomats get familiar with AI tools, regardless of their current proficiency.
Browse an index of Expo 2025 Osaka pavilion reviews—with a simple evaluation framework and links to detailed write-ups like Spain.
After fifteen years in diplomacy, I now work on how organizations actually change—and why most tech initiatives don't work. Director at Guidehouse, writing about transformation, technology adoption, and the messy human side of both. Subscribe to follow along.
I was about seven years into my career before I had a manager who did regular 1:1s with me, not the quarterly check-ins or annual reviews I was used to, but actual recurring conversations about work, about thinking, about how things were going. About four months in, we
DoW designated Anthropic a supply-chain risk. If you built workflows around Claude Code, here's how to protect your investment by running two AI tools against one shared vault.
The DoW designated Anthropic a supply-chain risk. If you built workflows around Claude Code, here's how to protect your investment with portable, vendor-agnostic AI setups.
Enterprise AI doesn't have to repeat the oldest mistake in IT. A three-layer model — shared data, standardized skills, and personalized agents — lets organizations amplify individual judgment instead of standardizing it away. The technology exists today.
Feedback built on aggregated impressions — "multiple people have raised concerns" — strips away everything that makes feedback useful. The research is clear: effective feedback is specific, actionable, and timely.
Most LinkedIn advice falls into one of two categories: tactical tricks that aim to game the system or vague encouragement to "post consistently and engage authentically." Neither helps you understand how the platform actually works, which is what you need to make good decisions about your own strategy.
For the reasoning behind this system, see Why Feedback Order Matters: A System for Document Review. The Three Tiers TierScopeAddressExamplesHigher Order (HOC)Whole documentFirstArgument, structure, purpose, audience fit, completenessMiddle Order (MOC)Sections/paragraphsSecondTransitions, evidence integration, tone, proportionalityLower Order (LOC)Sentences/wordsLastGrammar, spelling, word choice, formatting The core rule: Always work
Bad document feedback was the bane of my existence at the State Department. (We always talked about "happy to glad" changes). You've probably seen this happen. Someone sends a draft for review. The first reviewer opens it, spots a comma splice on page one, fixes it,
LinkedIn's algorithm has never been intuitive for users of other social media platforms: from its preference for once-a-day posting to how it values different forms of engagement. This got even more complicated in late 2025, when LinkedIn launched the 360Brew algorithm, a 150-billion parameter language
Every conversation with AI starts from scratch. This tutorial builds the system that fixes that — persistent context, project memory, and a personal vault that makes each session smarter than the last. Built for diplomats, analysts, and federal professionals who need more than a chatbot.
This guide walks you through setting up an AI agent with persistent memory, modular skills, tool integration, and task automation, without exposing endpoints to the internet. Time required: ~2 hours Difficulty: Comfortable creating folders and text files; some terminal use helpful but not required Prerequisites: macOS/Windows/Linux, Anthropic API
Two weeks ago, the Pentagon put Google's Gemini for Government on the desktops of 3 million military personnel, civilians, and contractors. If you have a CAC, you now have access to a frontier AI model cleared to handle CUI. Most people will poke at it a few times,
