The design process is dead. Here’s what’s replacing it. | Jenny Wen (head of design at Claude)
The design process is dead. Here’s what’s replacing it. | Jenny Wen (head of design at Claude)
The design process is dead. Here’s what’s replacing it. | Jenny Wen (head of design at Claude)
Jenny Wen, Design Lead for Claude at Anthropic and former Director of Design at Figma, joins the podcast to explore how AI is fundamentally reshaping the role, tools, and mindset of product designers.
Jenny argues that the traditional design process—centered on discovery, static mockups, and linear iteration—is obsolete in the AI era. Instead, design work has split into two tracks: supporting rapid engineering implementation and defining short-term (3–6 month) visions amid accelerating change. At Anthropic, designers spend far less time mocking and more time collaborating directly with engineers, using AI tools like Claude Code and Claude Cowork to prototype in code and ship faster. She emphasizes that while AI excels at generating options and accelerating execution, human judgment, taste, and accountability remain irreplaceable—especially when resolving trade-offs or upholding brand trust. Jenny’s move from director back to IC reflects her belief that deep technical fluency is essential for future leadership. She identifies three key designer archetypes for today’s landscape: strong generalists, deep specialists, and adaptable new graduates. Finally, she champions psychological safety through radical candor, values 'low-leverage' leadership acts that build trust, and highlights frameworks like legibility to navigate early-stage, high-uncertainty innovation.
00:00
00:00
The traditional design process is obsolete
04:26
04:26
The traditional design process is dead, and it was already dying before AI
09:15
09:15
Designers now have access to coding tools, can be more involved in implementation, work closely with engineers, and prototype in actual code instead of relying solely on engineers
12:33
12:33
AI use cases are discovered only through real-world usage, not planning
17:56
17:56
A few years ago, 60–70% of design work was mocking and prototyping; now it’s down to 30–40%, with engineering collaboration and shipping rising proportionally
18:45
18:45
Their team is fully using Claude
20:03
20:03
Figma still fills a good gap by allowing exploration of multiple design options and handling visual and interaction details
22:25
22:25
Both designers and engineers are struggling to keep up with the rapid pace of work due to new capabilities
24:22
24:22
Launch products as research previews, admitting flaws but promising to iterate based on user feedback
30:25
30:25
Humans must sign off on AI-made decisions due to liability, as in radiology
31:38
31:38
Chatbots and terminals won't disappear as they offer flexibility and a new way to interact with models
40:35
40:35
Receiving and accepting critical feedback regularly is a vulnerable exercise for designers
45:42
45:42
Claude is good at turning messy content into something nice
46:16
46:16
Deep specialists are top performers in narrow domains like AI-driven UX research or advanced interaction modeling
53:42
53:42
Claude is not yet hireable as a designer, though it's good at presenting initial ideas
54:46
54:46
Testing products, fixing bugs, and making anniversary cards are high-leverage leadership acts
1:00:37
1:00:37
Radical candor combines care and directness to build high-performing teams
1:06:58
1:06:58
Three patterns for spotting breakthrough potential: dive deeper, notice excitement, recognize top-one-percent founders
1:12:51
1:12:51
Used Claude Cowork to analyze personal notes and create a rubric for evaluating design craft
