daily design digest
Methods outlive tools. This hub archives briefings on design thinking, discovery, critique, research ops, and facilitation — written for people who deliver workshops and need them to land, not perform. UCD Coach favours plain language over framework theatre: what changed, who said it, and what to try in your next sprint. Use it to refresh your practice, prep a stakeholder conversation, or give juniors a reading list that respects their time.
Cursor and Claude Code speak the same Model Context Protocol, so an MCP server config you write once now works in both without changes. That's a genuinely useful shortcut for anyone juggling AI tools.
claudefa.st
A Medium piece on agency workflows describes teams restructuring around "human plus helper" pairs rather than full AI swaps. The productivity and conversion stats it cites come from the author's own case studies, not an independent study.
medium.com
An August 2026 post argues "graph engineering" — structuring context for AI agents — is becoming the differentiating skill behind better agent output. It's early-stage and mostly discussed in engineering circles, not yet a standard design method.
facebook.com
Reporting on multi-agent AI systems this summer found that most failures trace back to orchestration and process gaps, not the underlying models being incapable. The same lesson applies directly to how you structure your own design workflow.
rauljitechnologies.com
Creativeboom rounds up design trends practitioners are tired of this year — practitioner sentiment, not hard data. Useful as a gut-check before leaning on a trendy style.
creativeboom.com
A design trends piece makes the case that most 2026 "trends" are noise and only a handful actually change how people work. Worth borrowing that filter before you spend a week copying a moodboard.
think.design
Industry surveys show most teams using AI agents keep a human review step before anything ships — augmentation, not replacement. That review habit is a method worth building now.
moclaw.ai
A widely-cited 2026 graphic design trend report puts AI-generated design at the top of the list — but most featured work pairs AI output with heavy human editing, not raw generation. Neo-brutalism, bento grids and elastic typography round out the list.
digitalsynopsis.com
A widely-discussed 2026 round-up argues AI work has shifted from clever one-off prompts to repeatable, multi-step workflows, backed by framework releases like Google's ADK and LangGraph 1.0. It's a useful mindset shift for how you approach AI-assisted design tasks, not just for people building agents.
richardkovacs.dev
Practitioners are starting to talk about designing for AI agents as an audience, not just humans — how an agent reads your UI and what context it needs. It's early, but worth watching.
instagram.com
New usage data on Microsoft 365 Copilot shows nearly half of AI interactions now support analysis and problem-solving, not just content production. The implication for designers: critique and judgement skills matter more as execution gets automated.
mexicobusiness.news
A guide published this August lays out concrete patterns — triggers, approval steps, escalation rules — for keeping people in control before agentic AI tools take autonomous action. It reframes 'human-in-the-loop' as a design deliverable, not just an engineering principle.
airia.com
DesignRush's August 2026 award winners across website, app, and branding categories are a snapshot of what's currently rewarded as strong design work. It's marketing, not research — but it's a free, dated benchmark for portfolio critique.
newsfilecorp.com
A widely-watched trends breakdown argues the industry is shifting from single deliverables — a logo, a screen — to full working systems, and from big generalist agencies to smaller specialist teams. Neither shift is brand new, but it's accelerating.
youtube.com
One agency's documented AI workflow uses AI for research synthesis, competitor analysis and prototype generation, but keeps human judgement for every decision that matters. Researchers at Polytechnique Montréal found AI helps most in four specific spots: research, kick-starting ideas, generating alternatives, and prototype exploration.
eleken.co
A widely-shared piece on how designers actually work with AI in 2026 argues prompting is table stakes — the real skill is building a 'context layer' (markdown files, decision logs, reference libraries) that AI tools can draw on. It's a useful reframe if you feel behind on AI workflows.
medium.com
Nielsen Norman Group's long-standing guidance — that around 5 users surface most usability problems — remains the reference point methods writers keep returning to as AI speeds up other parts of the process.
nngroup.com
Microsoft's August 2026 Copilot Studio update pushes a shift from vague single prompts to structured, multi-step agent workflows with defined skills and evaluation. It's enterprise-focused, but the underlying method is worth borrowing.
youtube.com
A workflow doing the rounds pairs AI generation with a simple written ruleset: a design.md file defining your visual and content rules before you prompt anything. It's a small process fix that keeps AI output from drifting.
blog.mean.ceo
Jakob Nielsen's July roundup revisits a classic usability principle — constraints that prevent mistakes beat error messages that explain them — right as AI tools tempt teams to over- or under-guide users.
jakobnielsenphd.substack.com
