Durable roles still change. They just keep the human in the highest-value loop.

The strongest positions are usually the ones coordinating people, making tradeoffs under uncertainty, or operating in messy real-world environments where accountability cannot be delegated to a model.

Interpretation lens

Safety here means relative durability, not immunity. Use AI to remove admin drag so more of your time stays on trust, leadership, and edge-case judgment.

Roles with stronger durability in the current wave

More durable

People managers in changing orgs

Managers retain core accountability for organizational structure, conflict resolution, morale, and performance across hybrid human-agent teams.

Tasks under pressure

  • Status reporting
  • Meeting transcription summaries
  • Calendar balancing

Durable work

  • Org design
  • Conflict resolution
  • Talent development
  • Strategic calibration

Best adaptation moves

  • Run leaner systems with AI
  • Coach teams through workflow change
  • Instrument output quality & evals

More durable

Solution engineers and technical sellers

Complex architecture design, security review navigation, and trusted client relationships remain heavily human-dependent even as AI accelerates RFP drafting.

Tasks under pressure

  • Proposal templating
  • Demo environment prep
  • First-pass RFP answers

Durable work

  • Technical discovery
  • Objection handling
  • Client trust building
  • Enterprise governance alignment

Best adaptation moves

  • Use AI for deep account prep
  • Shorten customer follow-up loops
  • Own customer-specific system design

More durable

Field operators and compliance-heavy roles

Unpredictable physical environments, real-time safety decisions, and legal/regulatory accountability resist pure digital automation.

Tasks under pressure

  • Routine paperwork
  • Log entry documentation
  • Standard compliance forms

Durable work

  • On-site judgment
  • Safety decisions
  • Accountability in edge cases
  • Hardware intervention

Best adaptation moves

  • Use AI for field prep and reporting
  • Reduce admin overhead
  • Keep human sign-off close to the physical work

More durable

Product leads bridging users and systems

High leverage belongs to leaders who can identify problem spaces, make system-level tradeoffs, and govern human-in-the-loop workflows.

Tasks under pressure

  • Spec first drafts
  • Backlog triage
  • Release notes drafting

Durable work

  • Strategic prioritization
  • System-level tradeoffs
  • User empathy
  • Model capability boundaries

Best adaptation moves

  • Treat AI as operating leverage
  • Use rapid prototyping to shape product bets
  • Own eval design & human-in-the-loop guardrails

AI-proof jobs guides for durable, judgment-heavy work

Practical guides for people in jobs safest from AI who want to turn trust, leadership, and domain judgment into compounding leverage.

x402 Protocol: How to Charge AI Agents for Your MCP Tools and APIs

Oct 30, 2026 · AI for Coding

x402 Protocol: How to Charge AI Agents for Your MCP Tools and APIs

A practical walkthrough of x402 payments on Cloudflare, from a paid MCP tool on testnet to deciding whether real agent revenue is worth chasing yet.

Brian Weerasinghe · 45 minutes

How to Use Claude for Excel: Setup, Prompts, and Limits

Oct 3, 2026 · AI for Work

How to Use Claude for Excel: Setup, Prompts, and Limits

Install Anthropic's official add-in, run the prompts that save the most time, and know where it stops being safe to trust.

Brian Weerasinghe · 20 minutes

Supercharging Claude Code: Skills, Plugins, and MCP for Advanced Workflows

Sep 27, 2026 · AI for Coding

Supercharging Claude Code: Skills, Plugins, and MCP for Advanced Workflows

How to turn Claude Code from a smart autocomplete into a disciplined, secure, agentic platform

Brian Weerasinghe · 1 hour

How to Install Claude Code: Native Installer, Homebrew, npm, and WinGet

Sep 27, 2026 · AI for Coding

How to Install Claude Code: Native Installer, Homebrew, npm, and WinGet

Every install path for macOS, Linux, and Windows, plus how to verify it worked and log in.

Brian Weerasinghe · 10 minutes

How to Self-Host n8n: A Docker Compose Setup Guide

Sep 16, 2026 · AI for Work

How to Self-Host n8n: A Docker Compose Setup Guide

Run n8n self-hosted for free with Docker Compose and HTTPS, plus the community edition limits and backup steps most tutorials skip.

Brian Weerasinghe · 45 minutes

FreeFlow Setup Guide: A Free Wispr Flow Alternative for Mac

Sep 15, 2026 · AI for Work

FreeFlow Setup Guide: A Free Wispr Flow Alternative for Mac

Install the open source dictation app, add a free Groq API key, and start hold-to-talk voice typing with no subscription.

Brian Weerasinghe · 15 minutes

How to Install Codex CLI and Use It in a Real Project

Aug 21, 2026 · AI for Coding

How to Install Codex CLI and Use It in a Real Project

A practical setup path from npm install to your first AGENTS.md-driven task

Brian Weerasinghe · 30 minutes

What Is an MCP Server? A Beginner's Guide

Aug 10, 2026 · AI for Coding

What Is an MCP Server? A Beginner's Guide

How AI apps use MCP tools, resources, and prompts to work with external systems.

Brian Weerasinghe · 15 minutes

Claude Code Skills: How to Install, Create, and Use Them

Aug 9, 2026 · AI for Coding

Claude Code Skills: How to Install, Create, and Use Them

Turn repeated coding instructions into reusable skills that Claude can discover automatically or run as slash commands.

Brian Weerasinghe · 25 minutes

Claude Code Subagents: How to Set Up and Use Them

Aug 7, 2026 · AI for Coding

Claude Code Subagents: How to Set Up and Use Them

Give Claude Code isolated helpers that keep your context window clean and your long sessions from falling apart.

Brian Weerasinghe · 20 minutes

How to Fix Codex Usage Limits: 7 Token-Saving Methods, Verified

Aug 3, 2026 · AI for Coding

How to Fix Codex Usage Limits: 7 Token-Saving Methods, Verified

Seven popular Codex token-saving tricks, checked against benchmarks and OpenAI's own docs, so you know which ones actually move the needle.

Brian Weerasinghe · 25 minutes

How to Run Windows on a Mac With Parallels Desktop

Aug 2, 2026 · AI for Work

How to Run Windows on a Mac With Parallels Desktop

Set up Windows 11 on your Mac in about 15 minutes, pick the right edition, and save 45% with code PD45.

Brian Weerasinghe · 20 minutes

How to Save Tokens in Claude Code Using Graphify

Jun 28, 2026 · AI for Coding

How to Save Tokens in Claude Code Using Graphify

Stop burning your Claude usage limits on file search. Build a knowledge graph once, query it forever.

Brian Weerasinghe · 20 minutes

How to Use Lovable AI: A Complete Guide for Non-Technical Founders

Jun 26, 2026 · AI for Founders

How to Use Lovable AI: A Complete Guide for Non-Technical Founders

A beginner's walkthrough for building and publishing a real website or app with Lovable, without writing a single line of code.

Brian Weerasinghe · 45 minutes

Figma MCP Server Setup: Claude, Codex, Cursor, Copilot, and Antigravity

Jun 20, 2026 · AI for Coding

Figma MCP Server Setup: Claude, Codex, Cursor, Copilot, and Antigravity

Connect Figma's Dev Mode MCP server to every major coding agent and start turning real design data into code.

Brian Weerasinghe · 20 minutes

How to Humanize AI Text: 7 Editing Moves That Work

Jun 7, 2026 · AI for Work

How to Humanize AI Text: 7 Editing Moves That Work

Practical techniques to edit AI-generated content so it reads like a person wrote it and passes the detectors checking for it

Brian Weerasinghe · 20 minutes

How to Set Up Hermes Agent: The Self-Improving AI That Runs 24/7

May 31, 2026 · AI for Founders

How to Set Up Hermes Agent: The Self-Improving AI That Runs 24/7

Install Nous Research's open-source autonomous agent locally or on a VPS in under 20 minutes, complete with model setup, Telegram gateway, and code snippets

Brian Weerasinghe · 20 minutes

How to Get Claude Code Free Using OpenRouter's Free API

May 30, 2026 · AI for Coding

How to Get Claude Code Free Using OpenRouter's Free API

Route Claude Code through OpenRouter's free model tier and run a full agentic coding environment at zero cost - legally, permanently, and in under 10 minutes.

Brian Weerasinghe · 10 minutes

How to Use Claude Inside Microsoft 365

May 8, 2026 · AI for Work

How to Use Claude Inside Microsoft 365

Claude now works natively inside Excel, PowerPoint, Word, and Outlook. Here is how to get it running and what to actually do with it.

Brian Weerasinghe · 15 minutes

How to Get Started with Claude Code: A Beginner's Complete Guide

May 3, 2026 · AI for Coding

How to Get Started with Claude Code: A Beginner's Complete Guide

From installation to your first real coding session - everything you need to go from zero to productive with Anthropic's terminal-based AI coding agent.

Brian Weerasinghe · 30 minutes

Set up useful AI teammates with new ChatGPT workspace agents

May 15, 2024 · AI for Work

Set up useful AI teammates with new ChatGPT workspace agents

Turn generic AI into dedicated expert teammates by creating workspace-specific agents in ChatGPT.

Brian Weerasinghe · 20 minutes

Jobs safest from AI: common questions

What jobs are safest from AI?

In AETW's task-exposure scores, the safest jobs from AI are skilled trades (14/100) and product managers (28/100). Among office and field roles, the most durable are people managers in changing orgs; solution engineers and technical sellers; field operators and compliance-heavy roles; and product leads bridging users and systems. They rest on judgment, trust and accountability.

What makes a job AI-proof?

No job is fully AI-proof, but the durable ones share three traits: decisions made under uncertainty, accountability that can't be delegated to a model, and work in messy physical or human settings. AI still takes over the admin around these jobs, which frees more time for the parts that stay human.

Which industries are least exposed to AI?

The least exposed industries in AETW's industry matrix are Skilled Trades (14/100), Construction (18/100) and Education (42/100). Hands-on, on-site work is still hard for AI agents to take over.

Are any jobs completely safe from AI?

No. Safety here means relative durability, not immunity. Even the most durable roles are changing as AI removes routine paperwork, reporting and scheduling. The people who use AI to clear that work tend to gain the most.

Community builderCommunity builderCommunity builderCommunity builder
Read by engineers & leaders

The AI brief for builders, operators, and leaders

Follow the AI developments reshaping work and the world, with practical context for what to do next.

Free, no spam, unsubscribe anytime. By subscribing you agree to our Terms and Privacy (16+).