Food for Agile Thought #545: Real Life Agentic Chaos, Product Leadership & AI, AI Killed the Agile Industry, Assembly Line Comeback

TL; DR: Agentic Chaos — Food for Agile Thought #545

Welcome to the 545th edition of the Food for Agile Thought newsletter, shared with 35,577 peers. This week, Natalie Shapira et al. reveal how autonomous LLM agents leak information, spoof identities, and falsely report task completion when red-teamed in a live lab, a finding that sharpens the question Charlene LiLi raises with David Burkus: AI transformation fails when CEOs hand it off to IT because the real challenge is behavioral, not technical. April Dunford picks up the strategic thread, urging companies to rethink their positioning by forming a clear point of view about the future rather than chasing speed. Petra Wille - loomista echoes that theme in an interview with Jason Knight, arguing product leadership itself demands deliberate development, not just promotion. And while Peter Saddington declares AI has inverted every value of the Agile Manifesto, McKinsey doubles down on industrial thinking with an “AI assembly line” that decomposes knowledge work into standardized agent tasks.

Next, Ant Murphy reframes prioritization as a layered chain of decisions flowing from vision to outcomes, not a backlog exercise, while Petra Wille challenges product leaders to resist AI hype and take responsibility for shaping a future worth living in. Paweł Huryn offers a practical tool for that effort with PM Brain OS, an open-source second brain built on markdown and Claude Code. Yet building reliable AI systems remains elusive: Swarnendu Bhattacharya reports that 88% of AI agent projects fail because teams rely on prompts rather than deterministic constraints, and Andon Labs proved the point by giving four AI models their own radio stations only to watch them develop wild personalities while ignoring the business side entirely.

Lastly, Barry O'Reilly argues that AI reassembles tasks within jobs rather than replacing them, shifting value from routine friction to better judgment, a theme Seth Godin extends by urging people to use machines for leverage rather than competing against them. John Cutler reminds us that even defining teams honestly is hard because it exposes power structures that organizations prefer to ignore. Shreshta Shyamsundar and Anmol Jain push further, proposing an agentic P&L that replaces headcount with cognitive outcomes. Finally, Itamar Gilad challenges hyped AI PM archetypes in favor of one who improves all company functions, not just coding.

🎓 🇬🇧 The Claude Cowork Online Course — Available June 8–15 for $129

You have been prompting AI for months. The results are inconsistent, every conversation starts from zero, and the model forgets who you are. That is the ceiling of prompting.

The Claude Cowork Online Course teaches you to break through it: build Skills that encode your expertise, connect them to your tools, and assemble Agents who handle recurring work the way you would handle it yourself. No coding required.

What You Will Get:

✅ 8+ hours of self-paced video modules: Skills, Agents, delegation frameworks — ✅ Tested with a live BootCamp cohort (April 2026) — ✅ The A3 Framework: decide what to delegate and what to keep — ✅ Starter kit with folder structure, CLAUDE.md, and Skill templates — ✅ All texts, slides, prompts, graphics; you name it — ✅ Designed for the $20/month Pro plan — ✅ Lifetime access to the version you purchase — ✅ Claude Cowork Foundational Certificate.

👉 Please note: The course will be available for $129 from June 8 to 15, 2026! (After that, $199.) 👈

🎓 Join the Waitlist of the Course Now: Claude Cowork: Stop Prompting. Start Delegating. No Coding Required!

Did you miss the previous Food for Agile Thought issue 544?

🗞 Shall I notify you about articles like this one? Awesome! You can sign up here for the ‘Food for Agile Thought’ newsletter and join 35,000-plus subscribers.

🎓 Join Stefan in one of his upcoming training classes!

🏆 The Tip of the Week: Agentic Chaos

(via arXiv.org): Agents of Chaos

Natalie Shapira et al. red-teamed autonomous LLM agents in a live lab environment with persistent memory, email, and shell access. Over two weeks, twenty researchers uncovered eleven failure patterns, including unauthorized compliance, information leaks, identity spoofing, cross-agent propagation of unsafe behavior, and agents falsely reporting task completion while the system state told a different story.

Source: arXiv.org: Agents of Chaos

🎯 Product

Jason Knight and Petra Wille: Strong Product Leadership in the Age of AI

Petra Wille joins Jason Knight to discuss why product leadership requires deliberate development, not just promotion, and how AI is forcing leaders to rethink team structures, workflows, and decision-making beyond optimizing purely for speed.

Source: Strong Product Leadership in the Age of AI

Authors: Jason Knight and Petra Wille

April Dunford: Positioning in the Age of AI

April Dunford suggests that AI is forcing companies to rethink their positioning by developing a clear point of view about the future, positioning against real current competitors, and staying close to what customers actually think.

Source: Positioning in the Age of AI

Author: April Dunford

Ant Murphy: Prioritization Happens In Layers

Ant Murphy proposes that prioritization is not a backlog-level activity but a layered chain of decisions flowing from vision through strategy to outcomes, and that when it feels hard, you are likely operating at the wrong level.

Source: Prioritization Happens In Layers

Author: Ant Murphy

Petra Wille: The Future Is Not Yet Invented: The Product Leader’s Influence on the World We All Will Live in

Petra Wille challenges product leaders to resist AI hype, question premature layoffs driven by shareholder interests, and take intentional responsibility for shaping a future worth living in.

Source: The Future Is Not Yet Invented: The Product Leader’s Influence on the World We All Will Live in

Author: Petra Wille

🧠 Artificial Intelligence

📺 David Burkus and Charlene Li: The Reason Your AI Rollout Isn’t Working with Charlene Li

Charlene Li joins David Burkus to argue that AI transformation fails when CEOs outsource it to IT because of behavioral change. Strategic leadership, not tool adoption, is what actually drives organizational value from AI.

Source: 📺 The Reason Your AI Rollout Isn’t Working with Charlene Li

Authors: David Burkus and charlene li

Barry O’Reilly: How Is AI Reshaping Jobs Instead of Replacing Them?

Barry O’Reilly proposes that AI is not replacing jobs but reassembling tasks within them, shifting value from managing friction and routine work toward better judgment and decision-making, and redesigning how people contribute.

Source: How Is AI Reshaping Jobs Instead of Replacing Them?

Author: Barry O’Reilly

(via Andon Labs): We let four AIs run radio stations. Here’s what happened.

Andon Labs gave four AI models their own radio stations, including bank accounts and a profit mandate. Five months later, each developed a wildly different personality, and almost none figured out the business side.

Source: Andon Labs: We let four AIs run radio stations. Here’s what happened.

(via Medium): 88% of AI Agent Projects Never Reach Production. The Problem Is Not Your Model. It Is Your Harness.

Swarnendu Bhattacharya suggests that 88% of AI agent projects fail not because of model limitations but because teams rely on prompts instead of building deterministic system constraints, a discipline called harness engineering.

Source: Medium: 88% of AI Agent Projects Never Reach Production. The Problem Is Not Your Model. It Is Your Harness.

(via O’Reilly Media): The Agentic P&L: Beyond the Empire of Headcount

Shreshta Shyamsundar and Anmol Jain propose replacing the traditional headcount-driven P&L with an agentic model that measures departments by knowledge readiness, agent throughput, and cost per cognitive outcome.

Source: O’Reilly Media: The Agentic P&L: Beyond the Empire of Headcount

🖥 💯 🇬🇧 Claude Cowork BootCamp #2, June 10-July 2, 2026 — $249

You bought the Claude Pro subscription. You installed the desktop app. You pointed Cowork at a folder, watched it churn for 30 seconds, and got something that looked impressive but wasn’t quite useful. You tried again. Same result.

Most people who try Claude Cowork get stuck in the same place. They do not know which of their tasks are good candidates for automation. They do not know how to build Skills that survive a second use. They do not know where the current limits sit. So they keep treating Cowork like a chat tool, missing the point of an AI assistant who actually does the work.

The Claude Cowork BootCamp fixes that. In four hands-on sessions, you build working Skills and AI Agents during the sessions, not after. You leave with a compounding system, not a stack of prompts.

The Founding Cohort sold out in April 2026. Cohort #2 starts June 10, 2026.

A Note on Positioning: This BootCamp is built for knowledge workers who want to automate repetitive work with Claude Cowork. There is no Agile-specific content in the curriculum.

The Cowork BootCamp is in English. 🇬🇧

Learn more: 💯 🇬🇧 Claude Cowork BootCamp #2, June 10-July 2, 2026.

➿ Agile & Leadership

Peter Saddington: The Agile Industry is Defacto AI

Peter Saddington proposes that AI has inverted every value of the Agile Manifesto, turning documentation into the primary input for code generation and making the traditional Agile industry, its ceremonies, and its certifications obsolete.

Source: The Agile Industry is Defacto AI

Author: 𝙋𝙚𝙩𝙚𝙧 𝙎𝙖𝙙𝙙𝙞𝙣𝙜𝙩𝙤𝙣

Seth Godin: Value creation, bullshit jobs and the future of work

Seth Godin suggests that AI will eliminate jobs that feel pointless while creating new ones, and that resilience lies in using machines for leverage rather than competing against them at tasks they already do better.

Source: Value creation, bullshit jobs and the future of work

Author: Seth Godin

John Cutler: Why Defining Teams Is So Hard

John Cutler explains that defining teams is hard, not because it requires special expertise, but because honestly describing how work actually flows threatens existing narratives, incentives, and power structures that organizations prefer to leave undiscussed.

Source: Why Defining Teams Is So Hard

Author: John Cutler

📯 Dear Micromanager: Your Distrust Has a Job; It’s Just Not the One You’re Doing

Twenty years of agile coaching failed to fix the micromanager who meddles with every draft, every meeting, every decision. This article shows where their distrust stops damaging teams and starts producing the verification work AI adoption actually needs.

Welcome, Verification Architect!

Learn more: Dear Micromanager: Your Distrust Has a Job; It’s Just Not the One You’re Doing

🛠 Concepts, Practices, Tools & Measuring

Pawel Huryn: PM Brain OS: The Second Brain for Product Managers, Made of Markdown

Paweł Huryn introduces PM Brain OS, an open-source ‘second brain’ for product managers built entirely from markdown files and Claude Code, designed to preserve decision context, surface contradictions, and prevent the knowledge decay that kills most AI memory systems.

Source: PM Brain OS: The Second Brain for Product Managers, Made of Markdown

Author: Pawel Huryn

Itamar Gilad: 3 AI PM Archetypes + 1

Itamar Gilad examines three hyped AI product manager archetypes and proposes a fourth: the AI-empowered PM who uses AI to improve all company functions, not just coding or spec generation.

Source: 3 AI PM Archetypes + 1

Author: Itamar Gilad

(via McKinsey & Company): The AI assembly line: Strategic imperatives for CEOs

McKinsey proposes an “AI assembly line” where CEOs decompose cognitive work into standardized tasks executed by AI agents — essentially, Taylorism for knowledge work. The framework ignores decades of evidence on why industrial paradigms fail in complex, adaptive environments.

Source: McKinsey & Company: The AI assembly line: Strategic imperatives for CEOs

📅 Training Classes, Meetups & Events 2026

Upcoming classes and events:

• 🖥 💯 🇬🇧 May 28-June 25, 2026 — Live Virtual Cohort: AI for Agile BootCamp #7 (English)
• 🖥 💯 🇬🇧 June 8 — Self-Paced Online Course: Claude Cowork: Stop Prompting. Start Delegating. (English; Self-paced Online Course)
• 🖥 💯 🇬🇧 June 9 — Meetup: HoA 74: “CLAUDE.md” — The One File That Makes AI Remember How You Work (English)
• 🖥 💯 🇬🇧 June 10-July 2, 2026 — Live Virtual Cohort: Claude Cowork BootCamp #2 (English)
• 🖥 💯 🇬🇧 June 11 — Meetup: HoA 75: Token Economics (English)
• 🖥 🇩🇪 June 30-July 1, 2026 — Live Virtual Class: Professional Scrum Product Owner Training (PSPO I; German)
• 🖥 💯 🇬🇧 July 1 — Self-Paced Online Course: AI 4 Agile Course v3 — Master AI for Agile Practitioners (English; Self-paced Online Course)

👉 See all upcoming classes here

🗞️ The Previous Food for Agile Thought Edition

Read more: Food for Agile Thought 544: Knowledge Work Tools in 2026, Product Buy-In Trap, AI-Generated MVP Issues, Agentic Coding ROI.

📺 Join 6,000-plus Agile Peers on YouTube

Now available on the Age-of-Product YouTube channel to improve learning, for example, about AI’s Labor Market Impact:

• Stop Writing Prompts. Let AI Do It for You — Hack #01, AI4Agile Online Course v2.
• Socratic Prompting — Hack #10, AI4Agile Online Course v2.
• Check Your AI’s Plan Before — Hack #7, AI4Agile Online Course v2.
• From Product Requirements to Experiments to Learnings — Supported by Generative AI.
• Never Accept an LLM’s First Offer — Improve GenAI’s Usefulness w/ Feedback Loops and Challenges.

🎓 Do You Want to Read More Like This?

Well, then:

• 📰 Join 35,000 peers and sign up for my weekly newsletter
• 📅 Join 6,000-plus peers of the Hands-on Agile Meetup group
• 🐦 Follow me on Twitter and subscribe to my blog, Age of Product
• 💬 Alternatively, join 20,000-plus peers of the Slack team “Hands-on Agile” for free.

Food for Agile Thought 545: Real Life Agentic Chaos, Product Leadership & AI, AI Killed the Agile Industry, Assembly Line Comeback was first published on Age-of-Product.com.