Designers, 'should designers code' and AI-driven content pipelines

Karri Saarinen argues 'should designers code?' is the wrong question and urges focusing on problem definition. • He divides design into problem, concept, execution stages and emphasizes aligning stakeholders on the underlying problem before execution. • Tools shape behavior: starting in Figma or VS Code primes different approaches; coding-first workflows may skip slow, conceptual thinking. • Karri warns against losing 'slow thought' as execution gets faster with AI, and says prompt-writing skills (defining goals, constraints) become more valuable. • Commentary notes some domains (consumer/social) may benefit from fast prototyping; design trade-offs depend on context. • Separately, an AI pipeline illustrates end-to-end e-commerce creative generation: model-driven model photos, virtual try-on, group product images and video creation.
Key insights
Karri: ask better questions rather than demand designers code
Karri Saarinen contends the core question is what we expect designers to deliver in the AI era; he prescribes first validating the problem, then exploring concepts, then executing—arguing coding as a tool should not replace problem alignment and conceptual thinking.
Tools bias workflows and can erode slow conceptual work
Karri warns that using code or design tools as default starting points biases teams to skip 'slow thinking' and concept work; he suggests preserving time for problem definition and concept selection even as AI accelerates execution and prototyping.
AI enables full e-commerce creative pipelines from images to video
A separate case outlines an AI-driven pipeline that generates models, applies virtual try-on, creates strategic product image sets and produces short video content, demonstrating AI's capability to automate end-to-end creative production.