Bootcamp: how AI changed my design workflow in 2026

Published in Medium’s Bootcamp publication in June 2026, this piece by Avinash Kumar, a designer at WongDoody, describes eight stages of AI adoption across roughly two years of practice. It is not a tool review. It is a first-person account of how the relationship with AI changed what a working designer actually does — and what that change felt like from the inside.

The account is structured chronologically. Stage one starts where many designers started: using AI to fix English in case studies and client emails. Stage two is an honest admission — the author confused “Responsible AI” (an ethics framework) with a product, a common early mistake that illustrates how fast the terminology was moving. Stage three documents an unexpected creative phase: building AI-generated comic universes, learning to maintain character consistency across images using detailed prompt chains, and developing scene-by-scene visual storytelling.

Stage four is the FOMO phase — the experience of trying to keep up with every new release and failing. The author names this directly rather than glossing over it, which makes the piece useful for anyone currently stuck there.

The shift comes at stage five. The author stopped collecting tools and started treating AI as a thinking partner for design problems: generating edge cases in user flows, writing UX microcopy, exploring multiple directions quickly without the overhead of building each one. Stage six describes the moment when AI began holding project context across sessions — understanding the product, the tone, the user problems — which changed the nature of what could be delegated. Stage seven covers using AI as a second brain: consolidating research, design decisions, and notes into a system the AI could draw on, though the author also notes a concern about being unable to tell which ideas originated with them versus with the model.

Stage eight is vibe coding: designing functional landing pages and internal tools by describing what is needed in natural language, without writing production code. For the author, this removed the last excuse for not exploring enough.

The article is valuable for designers who are still in the tool-collection phase. It makes a clear argument that the meaningful shift is not about adding more AI features to a workflow, but about changing the mental model — from AI as accelerator to AI as collaborator. The author’s conclusion is deliberately understated: AI did not replace the designer’s role, it removed the reasons for not going further.

The writing is direct and free of hype. The author includes the awkward stages alongside the productive ones, which is more instructive than accounts that present AI adoption as a smooth upgrade.