AI has made building easier.
But building the right thing is still hard.
Ten articles every week on what is happening across product, research, design and engineering. Curated by Rory Madden to help teams decide what to build.
Explore the editions ↓The archive
Will Claude Code Ruin Our Team? by Justin JacksonAI coding is collapsing role boundaries across design, product, and engineering, so teams need new ways of collaborating that preserve shared judgment and respect instead of letting tool access turn cross-functional work into turf
Read edition ↗AI Doesn’t Reduce Work—It Intensifies It by Aruna Ranganathan and Xingqi Maggie AI can make teams faster, but without clear boundaries, it also expands workload, increases context switching, and makes it harder to switch off. Leaders need stronger norms around pace, focus, and re
Read edition ↗When companies launch transformational initiatives, they typically assign them to existing leaders who are already measured on current business KPIs. This creates an inevitable conflict: leaders will always prioritize the metrics they're evaluated on over new strategic projects.
Read edition ↗When companies empower teams without providing strategic context, a predictable pattern emerges: teams make logical decisions based on incomplete information while leadership sees these choices as misguided.
Read edition ↗The most effective visions paint vivid pictures of customer transformation through a systematic four-step approach:
Read edition ↗Most product visions end up as abstract statements that sound impressive but fail to inspire action or drive decision-making. To understand what makes a vision truly powerful, we analysed SpaceX's Mars mission video, which transforms the company from a transportation logistics pr
Read edition ↗Many organizations believe they have cross-functional teams when they actually have "teams of functions" - groups where each member still reports to separate functional managers, follows distinct processes, and adheres to rigid role boundaries with formal handovers between discip
Read edition ↗Level 1 is task execution. Level 2 is choosing the right solution and trade-offs. Level 3 is deciding what problem is worth solving. Level 4 is setting direction under uncertainty.
Read edition ↗AI has made competent analysis cheap. What it cannot give you is the messy, human truth: what is actually happening, which assumptions are wrong, and what people are not saying out loud.
Read edition ↗Empowered, autonomous teams can design and build better products because they are closer to the customers and can get quicker feedback.
Read edition ↗When anyone can generate a solid PRD, a decent design direction, or a passable analysis in minutes, the edge is no longer how well you produce. It is whether you are working on the right thing, and whether it actually moves an outcome.
Read edition ↗AI will build whatever you ask for, quickly and confidently. That is exactly why it is dangerous. It will not challenge the brief, spot the long-term trade-offs, or warn you when a “reasonable” choice breaks your architecture, your design system, or your product strategy.
Read edition ↗When products fail to deliver expected business value, most organisations respond by tightening control through more detailed business cases, approval layers, and coordinators.
Read edition ↗