The one-sentence summary

Success always starts with failure, and in a rapidly-changing world adaptive trial-and-error processes are best for tackling major issues.


  • Success always starts with failure, and here’s the evidence.
  • Out go leaders with a grand vision, experts with a detailed plan of action, and gurus with an infallible solution. In come those who can improvise rather than plan, working from the bottom up, and taking baby steps rather than great leaps forward. A ‘worm’s eye view’ is best.
  • When faced with complex problems we have all become accustomed to looking to our leaders to set a plan of action, but in fact the world has become far too unpredictable and profoundly complex for that to work. Instead we need to adapt.
  • Adaptive trial-and-error processes are better at tackling the major issues such as climate change, poverty and the financial crisis.
  • Philip Tetlock spent 20 years checking if the predictions of experts came true – most didn’t.
  • Big pictures and plans no longer work: “No plan survives first contact with the enemy.” (Prussian General)
  • Variation means set lots of ideas running – don’t nail your colours to just one project. Selection means dropping the stuff that doesn’t work. Decoupling means reducing the impact of failure by not linking things together (those setting up domino-toppling world record attempts now put gates in so that only individual sections can fall over before the main event).


  • This is a broad sweep of psychology, evolutionary biology, anthropology, physics and economics.
  • Evolution succeeds because it uses ‘works for now’ solutions rather than striving for the ultimate. Business can learn from this.
  • He likens corporations to species – companies suffer extinctions too.
  • Palchinsky’s Principles look good (in 1901 he reported to the Tsar that Russia’s coalmines were death traps):
  1. Seek out new ideas and try new things
  2. When trying something new, do it on a scale where failure is survivable
  3. Seek feedback and learn from mistakes as you go along
  • If a test is FUQed it may be because the issue cannot be dealt with by an experiment – Fundamentally Unidentified Questions can’t be answered.


  • The author is the self-styled ‘Undercover Economist’ and certainly knows his stuff, but it can veer off into areas like climate change that rather peter out.
  • He is billed as Britain’s Malcolm Gladwell. It’s certainly a good read: close, but no cigar.