Algorithms to Live By

Algorithms to Live By Summary

The Computer Science of Human Decisions

by Brian Christian and Tom Griffiths

  • 15 min read
  • Published 2016
  • 9 takeaways

Perfect information is a luxury good, and life keeps asking for boarding decisions. This is a sharp tour of what computer science can teach human judgment when time, memory, and patience refuse to expand.

What you'll learn
  • How to stop searching
  • Why schedules reveal values
  • The logic of useful clutter
  • When to explore new options
  • Why forecasts need humility

Key point 1

The crowded runway

A plane cannot circle forever while the control room waits for perfect information.

That is the useful nerve of Algorithms to Live By. Brian Christian, a writer with a programmer’s sense of limits, and Tom Griffiths, a cognitive scientist, take ideas from computer science and ask what they can teach ordinary human judgment.

Their strongest claim is bracingly practical. Many life problems are hard because time, attention, and memory are limited, so the best answer is often the one that handles those limits honestly. A perfect plan that arrives after the plane has landed is just theater.

The book does not say we are computers. It says computers have spent decades facing our least romantic problem: how to choose when every choice has a cost.

From here, the runway gets busier.

Key point 2

Stop when the sample has taught you enough

In 1960, Martin Gardner brought the secretary problem to a wide audience in Scientific American, and its answer still feels rude to human hope.

The classic version says you must interview candidates one by one. You can accept or reject each person, but you cannot go back. The best strategy is to look at roughly the first 37 percent, choose no one, and then pick the next candidate who beats everyone you have seen.

Good judgment often starts with a planned period of not choosing.

Christian and Griffiths use this as the book’s first great lesson in stopping. We like to treat searching as a sign of care. More apartments, more dates, more reviews, more tabs. Yet search has a price, and the price is not only money. It is missed chances, tired attention, and the slow comedy of becoming very informed about options that are already gone.

The control room learns something during the first stretch of scanning. It builds a standard. Then the work changes. The goal is no longer to know the whole sky. The goal is to land the best plane that appears after the learning window closes.

This matters because modern life sells us endless comparison as wisdom. The algorithm gives comparison an expiry date. It says that after enough sampling, more looking can become fear dressed as research.

The 37 percent rule is not a spell. It assumes a clean setup, a known number of options, and no second chances. Real life is messier and more forgiving. Still, the pattern is powerful. Set a learning phase. Respect it. Then act when something clears the bar.

Indecision has excellent manners. It still steals your seat.

Key takeaways

Key point 3

Your day breaks where the queue breaks

Key point 4

Keep the useful thing close

Key point 5

Explore while the price is still low

Key point 6

A humble forecast beats a loud one

Key point 7

When the protocol belongs to someone else

Key point 8

The instruments stay on

Key point 9

Try this

Continue reading the full book summary and unlock all remaining key takeaways.

Get full summary

About the author

Brian Christian and Tom Griffiths

Brian Christian is a writer and computer scientist known for translating technical ideas into humane, sharply observed prose. Tom Griffiths is a cognitive scientist and professor whose research connects computation, psychology, and human learning, making the pair unusually well suited to explain why our messy choices often have elegant structures underneath.

Related topics

Want to keep reading this summary?

Get full access to complete summaries and audio versions in one place.

Continue to onboarding

Related books

Keep learning with similar reads

Unlock full library

Frequently asked questions