Newsletter #65 - Frank Gehry Refused to Pretend

Frank Gehry Gives the Middle Finger

In October 2014, Frank Gehry stood at a press conference in Oviedo, Spain, accepting the Prince of Asturias Award for the Arts.

A journalist badgered him about his work being "showy" and "spectacle."

Gehry raised his middle finger.

When pressed for a verbal response, he gave one.

"In the world we live in, 98 percent of what gets built and designed today is pure shit. There's no sense of design nor respect for humanity or anything. They're bad buildings and that's it."

He continued.

"Every now and then, a small number of people do something special. We dedicate ourselves to our work. I work with clients who have respect for the art of architecture. Therefore, please don't ask questions as stupid as that one."

Gehry passed away at 95. He left behind buildings that reshaped skylines and a body of work that proved architecture could be art.

But he also left behind something harder to replicate.

A refusal to pretend.

Gehry's audacity came from knowing what he stood for. In a world where leaders round their edges, soften their opinions, and optimize for likability, there is something refreshing about someone who simply says what he believes.

You don't have to flip off reporters to lead with conviction. But you do have to know what hill you're willing to die on.

Zero-Downside Tasks

Effective delegation starts by letting yourself (and your assistant) fail in low-stakes ways.

The instinct is to delegate important things first. That's backwards. You wouldn't start learning to drive on the highway.

Build your delegation muscles with zero-downside tasks:

• Start with "time-only" risks → Choose tasks where the worst outcome is time spent by your assistant. Research summaries. First drafts. Meeting prep. Calendar audits. These are your practice ground. If the output misses, you lose an hour. You gain data on how to delegate better next time.
• Define minimum wins → Perfectionism kills delegation before it starts. Instead of asking for "the best option," request specific, achievable outcomes. "Find 3 interesting data points." "Outline 2 potential solutions." "Surface the top 5 candidates." Make success concrete. When the bar is visible, your assistant can clear it.
• Use 2-minute feedback → Scan quickly for value. Make a binary keep/revise decision. Give one specific direction for improvement. That's it. Lengthy reviews and detailed critiques slow the loop and train you to over-invest in tasks you delegated to save time in the first place.
• Iterate rapidly → When something misses the mark, resist the urge to take it back. Refine your instructions. Add context you assumed was obvious. Let your assistant try again with better guidance. Every revision sharpens the system.

As these low-stakes wins stack up, gradually increase task complexity.