Designing superintelligence: why OpenAI's design chief calls this the golden age for designers

Key points

Key takeaways from the Lenny's Podcast interview with OpenAI Head of Product Design Ian Silber (August 2026):

Designers are the unhappiest role in tech right now. Silber points to Lenny Rachitsky's tech-workforce sentiment survey, where designers and user researchers score the most overwhelmed, anxious and least optimistic on every dimension — driven by uncertainty over what a designer is now expected to be.

But he argues it is the best time in history to be a designer. Everyone is still extremely early, so no one should feel behind: someone starting today has a leg up on most people, with huge room to change how great products get made.

Productivity has not scaled the way engineering has. Engineers have 10x-ed or 100x-ed output with coding agents, yet the design process stays messy and fluid — you try things, most fail, and the traditional prototype-mock-test-research-iterate loop no longer fits the schedule.

AI is already an incredible product designer. Silber says it is genuinely great and accessible to everyone, though not yet the best at visual craft, typography or interaction design; humans still win on truly understanding user needs, inventing the new, and holding a point of view.

How OpenAI designs ChatGPT and Codex. For some surfaces the team obsesses — trying 100 things and shipping one — while for others (like the Codex "super app") they build in public, take big swings and learn fast, staying un-precious until the technology hardens.

Silber's mantra to his team: "just do less." Don't design it if you don't have to — reuse an existing component, extend a feature, or drop it entirely. Systems thinking and composable primitives now matter more than one-off screens.

Roles are blurring but not collapsing. PM, design and engineering increasingly overlap, yet Silber still believes distinct responsibilities add value; at a startup he would hire the best generalists, but at OpenAI's scale a well-rounded team of specialists wins.

What he hires for. No AI background required, but high curiosity, strong prototyping, systems thinking and a clear point of view — plus the humility to keep re-testing tools that failed weeks ago, because "the model we have today is the worst it will ever be."

Read more: Lenny's Newsletter · Dive Club