Who decides, and how? Matching decision stakes to decision roles Decisions stall when people are unclear about the role they’re meant to play. Naming both the type of decision and the role helps everything move faster.
This week's best things #114 A critical PHP flaw, fragile internet infrastructure, fear-based decisions, Wellcome’s bot-traffic, an ARG at the National Library of Scotland, voice transcription governance, pay-later messaging, literacy and LLMs, organising as speculative fiction, art-history benchmarks, and AI grant assessment.
This week's best things #110 An update on the impact of AI overviews on search traffic, the BBC is testing a new recommendations engine, a critique of wall label language, the Trussell Trust got a fully agent-made donation. Also: digital guides, museum-goer demographics, AI-written speeches, and naming colours.
This week's best things #108 Bayeux Tapestry interactive; writing as thinking, and what AI does to it; a book on newsletters; an AI confessions project; an HBR framework for team readiness; genius as social contagion; a (bad) piece dismissing the idea of enshittification; and some webinars and events.
A bad system will defeat a good person every time Across the cultural sector, digital practitioners are blaming themselves for hidden work that nobody planned or resourced. My research finds the same pattern everywhere and traces it to structural conditions that predate any individual.
This week's best things #106 An ARG woven through the National Library of Scotland, team taxonomies, algorithmic sameness in culture, a British Library image search, a 'Six Thinking Hats' GPT, sauna theatre, Meta's failed AI overhaul, LinkedIn's AI slop button, and some words we need
Departments of Doubt and Wisdom Councils Does your organisation have any structural mechanism for doubt at all? Somewhere that 'the plan' and core assumptions gets usefully challenged, by people who have no stake in defending them.
This week's best things #104 The Met launches semantic search, the impact of 'decision debt', AI is rewiring how we think, write and collaborate, brand debt liability, accessibility is a practice not a project. Plus equidistant pubs, 3D seat views, and World Cup data portraits.
This week's best things #103 Adaptive hierarchy, decision rights, eroding psychological safety, AI as environment, interface design, modern-work overload, IKEA reskilling, service design vs UX, public AI for cultural heritage, plus overmorrow, Criterion browsing, and No Clean Sheet.
This week's best things #102 Wifärt Gallery, a kerning tool, SMK's AI visitor guide, BFI's beta cinema site, Torchbox on user assumptions, No Web Without Women, digital trust, AI writing's shape, why innovation labs fail, the "illusion of effort" in AI-perceived work, and a wood-chopping simulator.