Keep pulling the thread on Jenny Wen.
The traditional design process, characterized by extensive research, discovery, and iteration cycles, is now considered "basically dead" due to the increased speed of engineering enabled by AI tools.
The time allocation for designers has shifted significantly, with mocking and prototyping decreasing from 60-70% of their time a few years ago to 30-40% currently.
Jenny Nguyen predicts that while chat interfaces will remain a primary way to interact with AI, many graphical user interfaces will eventually be generated by the models themselves rather than being hand-coded.
When hiring designers, Jenny Nguyen is currently focused on three key archetypes: "strong generalists" (block-shaped), "deep specialists" (deep T-shaped), and "craft new grads" who are humble, eager learners.
An internal Anthropic prototype called "Claude Studio" served as the origin for the skills framework and markdown-based instructions now used to direct Claude's behavior.
A designer's role now involves spending 30-40% of their time collaborating directly with engineers and taking on direct implementation and coding tasks.
Due to the rapid pace of AI development, design vision planning horizons have shrunk from multi-year (2, 5, or 10 years) to a much shorter timeframe of three to six months.
The non-deterministic nature of AI models makes it impossible to mock up all possible states, forcing product teams to test with live models and real user data to discover use cases and build effectively.
Current AI coding tools are described as "super linear," making them less effective for exploring a wide breadth of design options compared to a tool like Figma, which facilitates parallel exploration.
A current limitation of Claude's code generation capability is that it does not always correctly identify and use components from an established design system.
Anthropic strategically releases early-stage products like Claude Cowork as a "research preview" to manage user expectations about quality while committing to rapid, feedback-driven iteration.
Jenny Nguyen predicts that AI will improve its capabilities in traditionally human-centric skills such as taste, judgment, and design.