Article illustration 1

In the high-pressure world of defense technology, Palantir is renowned for its enigmatic culture and rigorous analytical prowess. Yet one of its core employee onboarding rituals seems jarringly out of place: a deep dive into Impro, a 1979 book on improvisational theater by Keith Johnstone. Former Palantir employee Nabeel Qureshi highlighted this in a reflection on his tenure, listing it alongside more predictable texts like The Looming Tower. At first glance, a guide for actors appears irrelevant to engineers building AI for national security. But as I discovered after immersing myself in Johnstone's work, Impro holds transformative insights for anyone in tech—from developers debugging code to CEOs steering startups. Its lessons on human dynamics cut to the heart of collaboration, innovation, and leadership in our industry.

Why Palantir Bet on Improv

Palantir’s choice isn’t random whimsy. The company operates in domains where stakes are life-or-death, decisions are rapid, and teams must adapt fluidly to ambiguity. Johnstone’s Impro, distilled from his work at London’s Royal Court Theatre, teaches precisely that: how to thrive in unstructured, high-stakes scenarios. As Mario Gabriele notes in The Generalist, Palantir’s recommendation reflects a deliberate strategy to cultivate what Johnstone calls "status experts"—individuals who intuitively navigate power dynamics and spontaneity. For tech leaders, this is gold. In sprint retrospectives or crisis meetings, understanding status transactions isn’t just useful; it’s critical for psychological safety and efficiency.

The Invisible Architecture of Status

Johnstone’s core thesis is radical: every human interaction is a status transaction, whether coding side-by-side or presenting to investors. We constantly play "high" or "low" status through micro-behaviors:
- High status: Holding eye contact, pausing mid-sentence ("errr..."), or occupying space expansively signals dominance or confidence.
- Low status: Breaking eye contact quickly, fidgeting, or minimizing physical presence conveys deference or uncertainty.

"The work was transformed," Johnstone writes of actors applying these principles. "Scenes became 'authentic'... Suddenly we understood that every inflection and movement implies a status."

In tech, these dynamics shape everything. A developer playing high status might dominate a stand-up, shutting down ideas; a low-status product manager might hesitate to flag a critical bug. But Johnstone argues true mastery lies in flexibility. Tech leaders can channel low status to spark psychological safety—like sitting on the floor during brainstorming sessions to signal openness, encouraging junior engineers to share wild ideas. Conversely, playing high status can project authority during investor pitches or outages, as seen in the decisive personas of figures like Jensen Huang. The key insight? Status isn’t fixed; it’s a tool. Ignore it, and you risk miscommunication. Harness it, and you amplify team cohesion.

Unlearning to Innovate: Crushing the "Education Barrier"

Johnstone rails against traditional education as a creativity killer, asserting it turns adults into "atrophied children." Tech, with its obsession with best practices and frameworks, often falls into this trap. Developers might rigidly adhere to SOLID principles, or data scientists default to textbook ML models, stifling originality. Johnstone’s antidote: unlearn to create. He cites Stanley Kubrick’s approach—"I don’t know. I’ve never seen anyone else light a film"—as emblematic of deriving solutions from first principles, not convention.

For engineers, this means questioning inherited dogma. Why use that legacy library? Could this algorithm be reinvented? Johnstone warns that viewing creativity as "self-expression" invites paralyzing self-critique. Instead, reframe it as "divine inspiration," where ideas exist independently of ego. In practice, this could look like:
- Guided improvisation exercises in sprint planning: "Guess" solutions to a problem without ownership, reducing fear of judgment.
- Red-teaming sessions where critiques must build on ideas ("Yes, and..."), not block them.

As Gabriele emphasizes, Palantir likely embraces this to combat groupthink in life-or-death software deployments, where unconventional thinking saves lives.

The "Yes, And" Imperative for Tech Teams

Improvisation thrives on accepting offers ("Yes") over blocking them ("No"). Johnstone illustrates:

ACTOR 1: Your name Smith?
ACTOR 2: Yes.
ACTOR 1: You’re the one who’s been mucking about with my wife then?
ACTOR 2: Very probably.

Blocking kills momentum; accepting builds collaborative narratives. In tech, this mirrors agile rituals. A developer dismissing a peer’s proposal ("That won’t scale") stalls innovation. Embracing it ("Yes, and we could containerize it") accelerates iteration. Johnstone notes life rewards "Yes"-sayers with adventure and "No"-sayers with safety—but in fast-evolving fields like AI, excessive caution breeds obsolescence. Leaders can foster a "Yes" culture by:
1. Rewarding experimentation in code reviews.
2. Framing failures as learning offers, not setbacks.
3. Using retrospectives to identify unconscious "blocks" in communication.

Beyond the Stage: A New Script for Tech Culture

Palantir’s endorsement of Impro isn’t a quirky anomaly; it’s a masterclass in leveraging human psychology for technical excellence. In an era where AI and cybersecurity demand unprecedented teamwork, Johnstone’s principles offer a blueprint:
- Status fluency prevents dominance hierarchies that derail pair programming.
- Unlearning unlocks the "divine inspiration" behind breakthroughs like retrieval-augmented generation or zero-trust security.
- "Yes, and" turns daily stand-ups into engines of innovation.

For developers and leaders alike, embracing improv isn’t about theater—it’s about rewriting the rules of engagement in a profession where the next line of code could change the world. As Gabriele’s exploration reveals, sometimes the most powerful tools aren’t in the tech stack, but in the unscripted spaces between us.

Source: Adapted from Mario Gabriele's analysis in The Generalist, October 2025.