# The Beautiful Mess The beautiful mess of cross-functional product development This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## A Day of Sensemaking After Starting a New Job DevFeed: [A Day of Sensemaking After Starting a New Job]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-09-11T02:16:25Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Notion](), [context](), [Tool]() Tags: [context](), [notion](), [product](), [workflows]() ### AI overview The author describes a day spent understanding a new product, its workflows, history, people, artifacts, and accumulated context. They use a subway-map metaphor and Notion tables to organize journeys, handoffs, data sources, and other materials, while questioning which information remains valid. ### Source excerpt I recently started a new job, which means I'm spending a lot of my time trying to figure out how everything works: the product, the workflows, the history, the people, the artifacts, and all the context that everyone else has accumulated over time. ## TBM 437: AI and the Recontextualization Tax DevFeed: [TBM 437: AI and the Recontextualization Tax]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-09-03T00:17:23Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Artificial Intelligence](), [context](), [meetings]() Tags: [ai](), [context](), [meetings]() ### AI overview The article argues that AI may reduce the organizational cost of recontextualizing complex information into simpler formats. It says AI can help with recontextualization and exception hunting when given sufficiently current raw context, allowing teams to work in varied ways while presenting information for different stakeholder audiences. ### Source excerpt Here's an area where I am optimistic about AI (for now, at least). ## TBM 437: Tokens, Hours, Points, and Other Curious Proxies DevFeed: [TBM 437: Tokens, Hours, Points, and Other Curious Proxies]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-08-17T23:04:04Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Artificial Intelligence](), [context]() Tags: [ai](), [comparisons](), [measurement](), [metrics]() ### AI overview An opinion article examines the difficulty of measuring the return on AI token usage. It compares tokens with hours and flow metrics, arguing that measurement depends on a broader theory of value and can favor short-term, easily measured use cases. ### Source excerpt Everyone is talking about "return on tokens." Vendors love it (as long as the news is good). ## TBM 435: 20 Unfiltered Operating Takes DevFeed: [TBM 435: 20 Unfiltered Operating Takes]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-08-08T00:31:35Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Framework](), [standard](), [Exception handling]() Tags: [advice](), [implement](), [simplify](), [strategy]() ### AI overview An opinionated set of operating advice argues that the North Star Framework is more useful as a teaching tool than as a rigid organizational practice. The article also cautions that vision-to-tactics pyramids can distract teams by turning labels and abstractions into substitutes for understanding real work. ### Source excerpt I had a friend reach out recently for advice. I started writing... ## TBM 434: How Maps Can Hide Problems DevFeed: [TBM 434: How Maps Can Hide Problems]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-08-01T09:13:07Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [context](), [structure]() Tags: [design](), [mapping](), [maps](), [organizations](), [ownership](), [problems](), [product](), [reporting](), [silos](), [structure](), [technology]() ### AI overview The article argues that organizational maps can conceal incoherence when strategy, structure, technology, incentives, goals, ownership, teams, and funding do not align. It contrasts coherent organizations, where context transfers across map layers, with incoherent organizations, where people must repeatedly reorient and translate. ### Source excerpt If what you are mapping is incoherent, don't fall in love with the map (or your personal ability to navigate with it). ## TBM 432: Bundling & Unbundling Capabilities (and AI) DevFeed: [TBM 432: Bundling & Unbundling Capabilities (and AI)]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-07-26T10:53:39Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Product Management](), [centralization](), [Development](), [Code]() Tags: [ai](), [centralization](), [code](), [development](), [product-management]() ### AI overview This commentary argues that capabilities once bundled into product management have separated into specialized roles and disciplines, while also being distributed across teams, tools, platforms, and external partners. It presents AI-related changes as one of several simultaneous patterns, including specialization, diffusion, centralization, integration, embedding, and rebundling. ### Source excerpt Early in my product career, I wore many hats. ## A Model for Understanding AI Success in Organizations DevFeed: [A Model for Understanding AI Success in Organizations]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-07-19T15:21:27Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Artificial Intelligence](), [context](), [domain](), [iteration](), [Risk]() Tags: [ai](), [context](), [domain](), [iteration](), [risk]() ### AI overview The article presents a multiplicative model for AI success based on technology understanding, problem understanding, practice evolution, and an organizational social contract. It argues that missing any essential factor can undermine the outcome, while resistance to AI mandates may reflect rational self-preservation and insufficient evidence. ### Source excerpt Here's a simple model for thinking about AI success in organizations. ## TBM 430: Incubate, Compound, Refinance, Liquidate DevFeed: [TBM 430: Incubate, Compound, Refinance, Liquidate]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-07-12T20:57:05Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Artificial Intelligence](), [Software](), [Code]() Tags: [ai](), [complexity](), [cost](), [coupling](), [product](), [value]() ### AI overview This opinion article combines software portfolio economics with AI, arguing that AI does not remove complexity, coupling, coordination, uncertainty, or decay. Instead, it can change the cost, speed, and feasibility of responding to them. The article uses a value-versus-carrying-cost model to describe software assets as incubate, compound, refinance, or liquidate cases. ### Source excerpt tl;dr: AI does not eliminate complexity, coupling, coordination, uncertainty, or decay. ## TBM 429: How To Thrive At Work When You Think Too Much DevFeed: [TBM 429: How To Thrive At Work When You Think Too Much]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-07-06T18:53:46Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [systems](), [help]() Tags: [problems](), [systems](), [understand](), [work]() ### AI overview An introduction to a proposed book about thriving at work when you feel misunderstood or tend to overthink. It examines how traits such as deep analysis, noticing inconsistencies, and questioning assumptions can be experienced as both strengths and flaws in workplace feedback. ### Source excerpt This post is important. ## How AI Amplifies Organizational Strengths and Dysfunction DevFeed: [How AI Amplifies Organizational Strengths and Dysfunction]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-07-03T20:15:44Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Artificial Intelligence](), [context]() Tags: [ai](), [behavior](), [dashboards](), [trust](), [workflow]() ### AI overview This commentary argues that AI can reduce workflow friction and support organizational improvement when teams have trust, openness, and a willingness to confront difficult truths. In organizations shaped by avoidance, political behavior, poor incentives, or the weaponization of metrics, AI may automate and intensify existing dysfunction instead. ### Source excerpt AI can be your scribe, your thought partner, data-stitcher, or your counselor. ## Why Product Development Bottlenecks Are Distributed and Context-Dependent DevFeed: [Why Product Development Bottlenecks Are Distributed and Context-Dependent]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-06-26T06:13:28Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [Development](), [systems](), [Loop Engineering](), [Large Language Model]() Tags: [ai](), [complexity](), [development](), [engineering](), [llm](), [product-development]() ### AI overview The article argues that product development is not a simple linear system with one bottleneck. It describes the work as a combination of sensing, deciding, learning, aligning, making, changing, supporting, and adapting, and says apparent bottlenecks are often temporary symptoms of interacting constraints. It acknowledges that LLMs can accelerate work in situations with abundant data, repetition, and situational awareness, but says this does not represent most cases. ### Source excerpt My goodness. ## TBM 425: AI and Agency DevFeed: [TBM 425: AI and Agency]() Original publisher: [Read original article]() Author: John Cutler Published: 2026-06-07T23:19:08Z Content type: opinion Language: en Sources: [The Beautiful Mess]() Topics: [AI Strategy](), [Artificial Intelligence](), [trust]() Tags: [agent](), [ai](), [ai-adoption](), [leadership](), [organizational](), [trust]() ### AI overview This commentary argues that companies' approaches to AI adoption can undermine employee agency, trust, and dignity. It presents resistance to mandates as a potentially rational response to organizational conditions and argues that employees should be able to learn how AI fits their work on their own terms. ### Source excerpt So you want your team to figure out how to use AI for the benefit of your company?