Lenny's Podcast: The design process is dead — here's what's replacing it

Jenny Wen is head of design at Anthropic. In this episode of Lenny’s Podcast, published March 1, 2026, she describes how AI has broken the classic “research → wireframe → test → iterate” sequence that defined UX and product design for the past two decades — and what the replacement looks like in practice.

The conversation is grounded in Wen’s current work: she leads design for Claude, which means she is simultaneously building a product that changes how design is done while managing a design team navigating those same changes.

Who it’s for. Product designers, UX designers, and design leaders at any level who feel the traditional process is under pressure but haven’t found a clear replacement framework. Also relevant for PMs and engineers who want to understand how design’s role is shifting and how to collaborate differently.

Key takeaways:

The traditional process broke because of time compression, not process flaws. Wireframes, low-fi mockups, and staged testing worked when generating high-fidelity options was expensive and slow. AI made rapid, high-quality exploration nearly instant, which means the gating function of each phase disappeared. The process isn’t wrong — its original constraint is gone.
The new role is decision-making, not artifact production. When AI can generate ten interface variants in the time it previously took to sketch one, the designer’s value shifts to judgment: which variant is right, why, and what needs to change. This is a fundamentally different skill from production craft, and it requires explicit practice.
Engineers are redesigning the designer’s role from the outside. Wen observes that engineers using AI coding tools now produce UIs that previously required dedicated design work. This isn’t a threat to be managed — it’s a shift in who does what. Designers who understand this dynamic can position themselves upstream, shaping intent before engineers build, rather than downstream, cleaning up after.
Context and taste cannot be automated. AI generates coherent-looking options with no understanding of the product, the user, the business constraint, or the brand. A designer who can inject that context — quickly, precisely, early — produces AI-assisted output that generic prompting cannot. Context quality has replaced craft execution as the primary differentiator.
The adjacent field is strategy, not aesthetics. Wen’s framing positions experienced designers closer to strategic decisions than visual production. Designers who move in that direction — influencing what gets built and why, not just how it looks — will find the transition manageable. Those who stay anchored to execution will face growing pressure.

Worth watching if you lead a design team and are trying to figure out what to stop doing, what to start doing, and how to explain that shift to stakeholders; or if you’re a mid-career designer who feels the tools changing faster than your mental model of the job.