Summary of "Why I Left Veritasium."

High-level summary

Petr (the narrator) explains why he left Veritasium after serving as its first full-time writer‑director from September 2020 to January 2025. He is grateful for the experience, has no bad blood with Derek or the team, and is launching his own channel, SciencePetr.

He recounts memorable production experiences, describes his training (physics undergrad + PhD in physics education research), and offers three core lessons he learned at Veritasium that shaped his approach to science communication.

He outlines what to expect from SciencePetr (AI safety plus broader science stories), asks for support (subscribe/share/Patreon), and states his editorial stance: rigorous, occasionally critical of science/technology, and radically optimistic about a better future.

Three core lessons (methodology and actionable steps)

1) Distinguish story from topic — tell stories to illuminate topics

Concept: A “topic” (e.g., geochemistry) is abstract; a “story” uses people, conflicts, and events to make the topic compelling and teachable.
Example: The scientist who measured Earth’s age via uranium–lead ratios — and the challenges (lead contamination → invention of clean rooms → fights over lead additives) — turns geochemistry into a motivating narrative.
Why it works: Stories create narrative hooks, let you teach science within a motivated framework, and connect to broader related narratives (e.g., Midgley → leaded gasoline → ozone/CFCs).
Actionable steps:
- When planning, find a human-centered narrative or conflict that naturally requires explaining the science.
- Use the story arc to sequence scientific explanations so viewers stay motivated.

2) Go deeper — pursue accuracy, original facts, and corrections

Concept: Dig into primary sources, correct errors, and present details audiences won’t find elsewhere.
Examples:
- Kodak/nuclear testing research: corrected reporting errors (photo film vs. x‑ray film; wheat husks vs. strawboard).
- Aircraft guidance animation: checked technical manuals and calculated which bit a cosmic ray would flip to ensure a 4‑second animation was accurate.
Inspiration: Robert Caro’s exhaustive research ethos — immerse yourself to find stories and facts others miss.
Actionable steps:
- Verify claims against primary sources and technical manuals when possible.
- Prioritize factual accuracy even for small details; correctness builds credibility.
- Aim for investigative depth so your work becomes the definitive artifact on a topic.

3) Go big and trust your gut — pick projects that matter to you

Concept: Algorithms matter, but projects you deeply care about tend to perform well or at least satisfy you creatively.
Editorial stance: Make ambitious, risky, or critical pieces even if they don’t promise maximum views. Be willing to make politically or ethically engaged work.
Examples: Critical videos on dual‑use figures like Thomas Midgley Jr., Fritz Haber, and Oppenheimer.
Actionable steps:
- Balance audience considerations (titles/thumbnails) with conviction about which stories need telling.
- Prioritize subjects that feel urgent or morally important, even if they risk virality.
- Use professional collaborators (animators, illustrators, editors) to realize ambitious visions.

Other themes and positions

Dual-use nature of science: Scientific knowledge is omni‑use — the same science can create enormous benefit or enormous harm (e.g., fertilizer vs. bombs; rocket fuel for space vs. ICBMs).
Criticism as love: If you care about science, you should also critique it to improve it and prevent harms.
Inspiration and norms: Carl Sagan is a primary model — care deeply, be willing to get political on existential risks (e.g., nuclear winter), and balance warning with radical optimism.
Editorial goals: Maintain high artistic and scientific standards (Sagan-level bar), be politically responsible, and be hopeful about a solarpunk, equitable future.

“Criticism as love.” “AI is the nukes of 2026.” — phrases reflecting Petr’s editorial urgency and tone.

What to expect from SciencePetr

Focus areas:
- AI safety and existential-risk discussions (including a 42‑minute collaboration with Palisade Research featuring Geoffrey Hinton).
- Broad science topics: physics, math, psychology, and historically or ethically important stories about science and technology.
- Critical, rigorous investigations alongside optimistic visions for a better future.
Production approach:
- High-quality animation and editing.
- Careful research and expert interviews.
Call to action:
- Subscribe, share, and consider supporting via Patreon to sustain high production values.

Context and background

Role and timeline: Petr was the first full-time writer‑director hired by Derek (Veritasium) and worked there roughly September 2020–January 2025.
Channel growth: Joined Veritasium at ~7M subscribers; left when it was ~17M. The team scaled from a few people to over 20 full‑time staff.
Training: Physics undergraduate degree and a PhD in physics education research (same supervisor as Derek). He delayed launching his own channel to learn and train at Veritasium.
Anecdotes and production highlights:
- Helicopter shoot with Adam Savage flinging pennies at Derek.
- Japan’s huge rainfall simulator.
- A 28‑hour iron‑smelting run to make katana steel.
- Meticulous animation and technical-verification work.

Speakers and sources featured

Petr (narrator; former Veritasium writer‑director; founder of SciencePetr)
Derek Muller (Derek; Veritasium founder/host)
Adam Savage
Emily (colleague; left to start her own channel)
Casper, Gregor, Henry, Sully, Peter, Fabio, Jakob (Veritasium colleagues)
Robert Caro (research influence)
Thomas Midgley Jr., Fritz Haber, J. Robert Oppenheimer (subjects of critical videos)
Carl Sagan (primary inspiration)
Geoffrey Hinton (interviewee in Petr’s AI video)
Palisade Research (collaborator on the AI video)
(Also referenced: an unnamed 1990s senator in connection to Kodak/nuclear testing reporting)