The design process is dead. Here’s what’s replacing it. | Jenny Wen (head of design at Claude)

Episode Summary

Executive Summary: Jenny Nguyen argues that AI is rapidly transforming design from a process centered on research, mocks, and handoff into one focused on rapid execution, prototyping in code, and guiding teams toward a clear direction. Designers must become more technical, more adaptive, and more collaborative with engineers while still providing taste, judgment, and accountability. She also outlines how Anthropic’s design team works, hires, and maintains quality in a fast-moving frontier environment.

Main Topics: AI is reshaping the design process (Priority: 5/5): Jenny says the classic diverge-converge design process is effectively obsolete in AI-era product development because engineering can now ship too quickly for designers to sit upstream and gatekeep mocks. Design work is splitting into execution and direction (Priority: 5/5): She describes two increasingly distinct modes: supporting fast implementation with engineers, and creating short-horizon vision/direction through prototypes rather than long-term decks. Designers must work more directly with code and engineers (Priority: 5/5): A larger share of design time is now spent jamming with engineers, polishing implementations, and prototyping in actual code; designers are increasingly part of the build loop. Figma still matters, but as part of a broader toolkit (Priority: 4/5): Jenny argues that Figma remains valuable for exploring many options and fine visual/interaction details, while coding tools are better for implementation and iteration. Trust and quality can be built through speed and iteration (Priority: 4/5): She explains that shipping early research previews can preserve trust if teams respond quickly to feedback and visibly improve the product over time. Hiring for frontier design requires new archetypes (Priority: 4/5): She looks for resilience plus three profiles: strong generalists, deep specialists, and exceptional new grads who are humble, fast learners, and eager builders. Management still matters, but must be closer to the work (Priority: 3/5): Jenny believes design managers remain useful, but the best managers now combine direction-setting with people management and should understand the current tool-driven workflow firsthand.

Key Arguments: Classic design workflows are losing relevance because engineers can create usable prototypes too quickly for designers to fully own upstream mockup-heavy work. AI products are non-deterministic, so designers need to test real implementations rather than rely only on polished mocks or clickable prototypes. Designers still provide crucial judgment: deciding what to build, what matters, and how to keep products coherent across many fast-moving contributions. Figma remains essential for exploring multiple directions and fine-grained interaction design, even as code-based tools become more central. Shipping early can be a strength if teams clearly label products as previews, iterate rapidly, and demonstrate that feedback is actually being used. The most valuable designers in this era are resilient, adaptable, and either unusually broad, exceptionally deep in one area, or early-career but highly capable and curious. Managers should not be purely administrative; they should understand the tools and workflows enough to direct teams effectively in the new design landscape.

Data Points: Historical time spent mocking/prototyping: 60 to 70% - Jenny said a few years ago most of a designer’s work was mocking and prototyping. Current time spent mocking/prototyping: 30 to 40% - She estimates the mocking/prototyping share has dropped significantly in her current workflow. Vision horizon in the past: 2, 5, or 10 years - She contrasted older long-range design visions with today’s shorter planning windows. Current vision horizon: 3 to 6 months - Jenny said visions are now often short-horizon prototypes or directions rather than long-term polished narratives. Anthropic co-work build time: 10 days - She noted the externally shippable version of co-work was built from an internal protoype to launch-ready product in about 10 days. Anthropic internal organizational size she managed at Figma: 12 to 15 designers - Jenny referenced the scale of her prior design org at Figma. Estimated current design split: 30 to 40% mocking, 30 to 40% engineer pairing, remaining slice implementation - She described how her day is now divided across design, collaboration, and hands-on shipping. Mercury customer count mentioned in sponsor read: 300,000+ entrepreneurs - Sponsor copy for Mercury cited its user base.

Pivotal Quotes: "Trust the process. That's basically dead." — Jenny Nguyen: Her core thesis on the outdated nature of classic design workflows in the AI era. "You as a designer actually like do not have the time to make these beautiful mocks anymore." — Jenny Nguyen: She explained why designers are moving away from mock-heavy work toward execution support. "At the end of the day, someone has to decide what is actually going to get built and what actually matters." — Jenny Nguyen: Her view on enduring human responsibility in AI-assisted product development.

Implications: Design roles are becoming more technical, iterative, and execution-oriented. Future designers will need comfort with code, ambiguity, fast feedback loops, and strong judgment—while managers must stay close to the work to remain effective.