Stefan Wolpers

AI on Top of a Dysfunctional System: The Product Backlog

by Stefan Wolpers | 2026-09-20 Featured Agile and ScrumAgile Transition

TL;DR: Polished Artifacts, Unchanged Decisions, Or Ten Backlog Anti-Patterns AI Makes Worse

Add AI to a Product Backlog process that already struggles, and everything seems to improve within an afternoon. The problem is that polishing artifacts with AI doesn’t fix the root cause: the basis for the team’s decisions doesn’t change; AI only removes the visible discomfort that used to signal something was broken, along with some of the pressure to fix it. AI applied to a dysfunctional system – here, the Product Backlog process – makes the dysfunction look like progress.

This is the first article of a new series that explains the uselessness of bolting AI onto a dysfunctional system. It explains why generative AI worsens ten Product Backlog anti-patterns, how polished AI artifacts can hide missing evidence and authority, and how teams can check whether their process is fit for AI.

The example anti-patterns accelerated by AI are from my Scrum Anti-Patterns Guide book.

AI on Top of a Dysfunctional System: 10 Product Backlog Anti-Patterns AI Makes Worse – by Stefan Wolpers of Age-of-Product.com

Thesis: Adding AI onto a dysfunctional system, for example, the Product Backlog, makes ten typical backlog anti-patterns worse, because polished AI artifacts hide missing customer evidence, decision authority, and feedback; the article explains these mechanisms, their costs, and how teams test whether their process is ready for AI.

Food for Agile Thought 562: Slop Grenades, AI Didn’t Make Delivery Free, Pacing the Frontier, Handling Disruptive Change

by Stefan Wolpers | 2026-09-18 News

TL; DR: Slop Grenades — Food for Agile Thought #562

Welcome to the 562nd edition of the Food for Agile Thought newsletter, shared with 35,321 peers. This week, Tobi Lütke tells Shane Parrish that AI makes judgment more valuable because machines cannot take responsibility. Teresa Torres, Petra Wille, and Josh Elman add that AI cheapens features and demos, yet working products still demand judgment. Barry O’Reilly makes that judgment auditable before deployment through baselines and accountable owners. Also, Dario Amodei applies that accountability to the labs with third-party evaluators and speed limits, while Dwarkesh Patel’s guests expect expert-beating AI within ten years.

Next, Nigel Thurlow argues that we turned Lean into belts and templates, although Toyota wanted a better management system. Refinement repeats that mistake: Maarten Dalmijn untangles its four parallel conversations, and Mike Cohn warns that polished AI stories hide gaps. AI adoption needs the same management focus: Aaron De Smet ties it to trust, Christiaan Verwijs and Barry Overeem recruit volunteers for Reflexive AI, Addy Osmani explains how unsupervised agents turn legacy code into technical debt factories.

Lastly, Cedric Chin argues that structural constraints, not stubbornness, kill incumbents, and Elena Verna catalogs the sales habits that smother PLG. Matthew Hodgson applies that logic to AI, where functional silos swallow productivity gains unless leaders fix the operating model before buying tools. Also, Eliezer Yudkowsky explains why models ignore instructions: a polite ambassador talks while a doer chases rewards. Finally, Marily Nika plans for that gap with Minimum Viable Quality: rank failure modes and design for recovery.

Food for Agile Thought 562: Slop Grenades, AI Didn't Make Delivery Free, Pacing the Frontier, Disruptive Change - Age-of-Product.com

The AI Workflow Inventory: Can Your Team Name the Work It Already Runs With AI?

by Stefan Wolpers | 2026-09-13 Agile and ScrumAgile Transition

TL;DR: The AI Workflow Inventory Finalizes the A3 Delegation System

You probably know your own AI shortcuts: the colleague who drafts the stakeholder update, the nightly job somebody set up before they left, the interview notes that go through a model on demand. However, I want to challenge you: that perceived knowledge can create false confidence, because it feels like knowing the team’s way of working with AI, when it is still only a diffuse understanding of the practice. Ask the team to combine those individual accounts into one list, and you may discover how little of the whole anyone can see. The AI Workflow Inventory is the artifact for that list: one row per recurring task, refined into provisional task classes to prepare the team’s next AI delegation decisions.

This is another post from the series on building the A3 Delegation System in public. (I gladly answer all questions you have, and yes, I am considering creating an A3 Delegation application.)

The AI Workflow Inventory of the AI Delegation System: Can Your Team Name the Work It Already Runs With AI? - Age-of-Product.com

Thesis: The AI Workflow Inventory is a one-page canvas listing every recurring task a team already runs with AI, grouped into provisional task classes. It is A3 Delegation System’s seventh artifact. This article explains how to build it in one 60-minute session.

Food for Agile Thought 561: Token Bills, Exhausting Change, Cracks in the AI Thesis, Automating Broken Systems

by Stefan Wolpers | 2026-09-11 News

TL; DR: Token Bills — Food for Agile Thought #561

Welcome to the 561st edition of the Food for Agile Thought newsletter, shared with 35,328 peers. This week, James Shore wants leaders to justify $21,000 monthly token bills per heavy user by measuring approaches, never people. Julie Zhuo adds that fear-driven mandates produce transformation theater, and Kate Leto argues AI exposes old leadership gaps that Petra Wille and Teresa Torres say individuals cannot fix. Brooke Weddle, Deepak Mahadevan, Richard Steele, and Tom Welchman tie AI value to operating-model redesign, while Brett Queener sees jobs collapsing into one application.

Next, Afonso Franco argues AI agents, your product’s second user, bypass the UI and break seat pricing, and Kyle Poyar shows how Notion, Rippling, and Profound survive almost weekly launches by separating shipping from announcing. Nigel Thurlow sees agents running bureaucracy at machine speed, and Richard Kasperowski remains the bottleneck of his Scrum team of six agents. David Pereira watches AI make unchecked assumptions comfortable, while OpenAI’s ChatGPT Work guide keeps decisions with humans.

Lastly, Ara Kharazian finds AI spend per employee at top firms fell nearly 10% in August as token prices dropped 41%, and Gergely Orosz reports Uber, Pinterest, and AT&T cut bills with open models. Attackers enjoy the same discount: Anthropic’s threat report shows agent swarms letting lone operators match state hackers, raising the stakes for Meta’s Muse agent booking travel and negotiating for you. Finally, Paul Iusztin warns weak plans make cheap models expensive.

Food for Agile Thought 561: Token Bills, Exhausting Change, Cracks in the AI Thesis, Automating Broken Systems - Age-of-Product.com

AI Transformations And Agile Transformations Rhyme

by Stefan Wolpers | 2026-09-06 Agile and ScrumAgile Transition

TL;DR: AI Transformations and Agile Transformations Rhyme

AI adoption seems to be scaling: 37% of respondents in McKinsey’s 2026 survey report an EBIT effect from AI, and Gartner finds that 22% of organizations have scaled it across business units. Now, Agile practitioners have seen this combination before, as AI transformations and Agile transformations rhyme. There are five classic failure patterns from the Agile transformation adventures that are back under new names: from mandates from above to licenses mistaken for training to greenfield showcases to parachuted consultants to promised payroll savings dressed up as strategy. They share one condition: organizations make AI decisions at organizational scale without leaving inspectable evidence at the workflow level in the trenches. And for good measure, let us throw in ignoring culture and excluding most of the organization’s people in the process.

The A3 Delegation System is not an AI-transformation method. It is a deliberately low-tech, paper-level discipline that lets a team produce that evidence in hours, and this article spends as much time on where the system stops as on what it does. If you are a Scrum Master, Product Owner or Manager, an Agile Coach, or a Project Manager who has to argue upward about AI, the last section gives you one thing to do this week.

AI Transformations And Agile Transformations Rhyme: The A3 Delegation System Is The Low-Tech Countermeasure – Age-of-Product.com

Food for Agile Thought 560: The Hugging Face Controversy, Evals for Product Teams, Canvas for Experiments, Skill Decay

by Stefan Wolpers | 2026-09-04 News

TL; DR: The Hugging Face Controversy — Food for Agile Thought #560

Welcome to the 560th edition of the Food for Agile Thought newsletter, shared with 35,342 peers. This week, Dwarkesh Patel and Ajeya Cotra examine AI agents coordinating, cheating, and hiding evidence, while Zvi Mowshowitz treats those behaviors as a warning against complacency in the Hugging Face controversy. Teresa Torres brings the response down to practice with AI evals, while Ethan Mollick keeps human judgment in place for consequential choices. Jane Fulton Suri reminds teams that insight grows through observation and co-discovery, and Nigel Thurlow shows why slack time gives people room for exactly that work.

Next, Benedict Evans argues that easier AI tool-building still leaves product managers with the harder job of finding the right problem. At the same time, Seema Amble maps where vertical AI can beat incumbents. Latent Space and Artificial Analysis temper agentic progress with rising costs, uneven gains, and hallucinations, as GPT-6 and Fable 5.1 become available. Afonso Franco shifts attention to the status signals that shape culture, as Addy Osmani warns that unsupervised outsourcing execution can quietly erode the judgment and repetition that build expertise. (The A3 Delegation provides a remedy here; see below.)

Lastly, Paweł Huryn shows how AI agents can build SaaS products without coding, making engineering literacy the key skill. Yanli Liu extends that idea by turning books and frameworks into reusable agent skills. Molly Stovold and Braden Kelley both tighten execution through fixed constraints, learning, and early kill decisions. Finally, Dan Luu offers a useful warning: confidence and bold claims mean little when the evidence does not hold up.