How to Succeed at Failing, Part 4: Extreme Resiliency (Update)

Episode Summary

Executive Summary: This episode argues that failure can be a source of resilience and innovation, using TB research, premortems, startup strategy, the Museum of Failure, and a college course on failure to show how anticipating, examining, and normalizing mistakes can improve outcomes. It contrasts productive failure-learning with paralyzing fear of failure, especially in high-stakes settings like medicine, engineering, and entrepreneurship.

Main Topics: Adaptive mistranslation in TB research: Physician-scientist Babak Javed found that bacteria can sometimes benefit from a moderate level of protein-making errors, a counterintuitive mechanism he called adaptive mistranslation. Premortems as a decision-making tool: Cognitive psychologist Gary Klein explains how imagining a project has already failed helps teams surface hidden risks, improve plans, and create candor, especially by empowering junior voices. Startup risk, resilience, and limits of premortems: Will Coleman of Alto argues that startups need focus on building, not dwelling on failure, though the company’s later struggles illustrate the instability of early-stage business models. The Museum of Failure and learning from bad outcomes: Samuel West’s museum showcases failed products and inventions to help people laugh, reflect, and learn from innovation gone wrong rather than only celebrating success. Teaching failure in higher education: Teresa McPhail’s 'Failure 101' reframes failure for engineering students as normal, subjective, and tied to uncertainty, anxiety, and even death, with the goal of reducing shame and building resilience. Institutional incentives that hide failure: Amy Edmondson and Roy Shalem discuss how academia and markets under-share null results and R&D dead ends, wasting knowledge that could reduce duplicate mistakes and speed discovery.

Key Arguments: Moderate errors can sometimes strengthen biological systems; in TB, adaptive mistranslation helps bacteria innovate until error levels become lethal. Premortems work because they remove uncertainty about whether a plan will fail, making it psychologically easier to identify risks honestly. Organizations often say they want insights, but they actually resist changes that those insights would require; harmony can suppress innovation. Startups may find premortems too negative if they are already operating under extreme uncertainty and high failure odds. Failure is a better teacher than success because it is more authentic, more diagnostic, and closer to how humans actually learn through trial and error. Academic and corporate systems lose valuable knowledge when failed experiments, null results, and dead-end R&D are not shared or monetized. Teaching students to accept failure can reduce anxiety and help them understand that setbacks are universal, not personal proof of inadequacy.

Data Points: TB annual deaths: more than 1 million people each year - The episode emphasizes TB as a major global killer, especially in the developing world. TB position among infectious diseases: #1 for the last 20–30 years - TB is described as the world’s deadliest infectious disease over recent decades. Javed’s experimental failure streak: over a year and a half without a single experiment working - Illustrates how failure is routine in scientific research. Premortem time horizon: six months to a year - Klein’s exercise asks teams to imagine the project failing in that future window. Alto revenue shock during COVID: 95% lost in a day - Coleman describes how the pandemic hit the ride-share startup’s business. COVID-era product contribution: many successful products became 20–30% of revenue - Coleman says incremental pandemic-driven changes made the business more robust. Museum of Failure collection size: more than 150 failures - West’s traveling museum displays failed products and inventions. Failure 101 completion threshold: A or F only - McPhail’s course uses a pass/fail-style grading system with only A for passing or F for failing. Students who failed the class: 2 - McPhail says only two students have ever failed the course. Premortem early sponsor objection: absolutely no way - The Air Force sponsor initially rejected the exercise as too depressing. R&D patent race example: 1876 - Bell and Gray filed telephone patents on the same day in 1876 as an example of winner-take-all innovation competition.

Pivotal Quotes: "TB, for the last 20, 30 years, has been the number one infectious disease killer in the world." — Stephen Dubner: Introduces the global health stakes of the episode. "Sometimes these errors in the right context and in the right degree can actually be good for the bug." — Babak Javed: Explains the idea behind adaptive mistranslation in TB bacteria. "A harmonious decision is a terrible idea because that means that everybody has a veto." — Gary Klein: Argues that consensus can block innovation and honest disagreement.

Implications: The episode suggests that institutions should design for intelligent failure: share bad results, run premortems, and teach people to distinguish useful setbacks from catastrophic ones. Doing so can improve science, business, education, and public health.