Algorithms to Live By
The Computer Science of Human Decisions
Brian Christian and Tom Griffiths · 2016 · Psychology
13 min · 12 chaptersNarrated in your voice
What it's about
Brian Christian and Tom Griffiths argue that many everyday dilemmas—when to stop searching, what to keep, how to schedule a day, when to trust a hunch—are versions of problems computer scientists have already solved. By translating results from optimal stopping theory, statistics, and game theory into plain language, the book shows that indecision is often not a personal failing but a hard computational problem, and that formal methods can tell us when 'good enough' actually is good enough.
Reading plan
Pick a finish date and Genius lays out the days — the plan shows today's target and keeps you honest.
Read together
Start a circle and share the code — everyone sees everyone's honest place in the book. Accountability, not leaderboards.
Who it's for
- A professional drowning in to-do lists who wants a principled way to decide what to do first - A reader curious about computer science who wants to see its ideas applied outside a lab - Anyone facing a big irreversible choice, like hiring, dating, or apartment hunting, who wants to know how long to keep looking
What’s inside
- 1Optimal Stopping1 min
The right way to decide when to stop searching and commit is to reject a fixed fraction of your options outright, then accept the next one that beats everything you have seen so far.
- 2Explore versus Exploit1 min
Whether to try something new or stick with a known favorite should depend mathematically on how much time you have left to enjoy the payoff, not on curiosity or habit alone.
- 3Sorting1 min
Organizing something fully is often a waste of effort, because the cost of sorting can exceed the cost of just searching when you need it.
- 4Caching1 min
What you keep physically or mentally close at hand should be decided by how recently you used it, not by how much you feel attached to it or fear needing it later.
- 5Scheduling1 min
A to-do list should be ordered by an explicit rule tied to deadlines or task length, not by which task feels most urgent or most dreaded.
- 6Bayesian Predictions1 min
How you should predict the future length or size of something depends on what kind of thing it is, and blindly extrapolating from current data without the right underlying pattern in mind leads to systematically bad guesses.
- 7Overfitting1 min
Adding more factors, more precision, or more effort to a decision can make the outcome worse, so deliberately simplifying is often the smarter strategy.
- 8Relaxation1 min
When a problem is too hard to solve exactly, deliberately ignoring one or more of its real constraints to solve an easier version first, then adjusting, beats either forcing an exact answer or giving up.
- 9Randomness and Game Theory1 min
Deliberately injecting randomness into a decision, and recognizing that trying to perfectly outthink other people's strategies is itself computationally hopeless, are both rational responses to a certain kind of complexity, not admissions of failure.
About Brian Christian and Tom Griffiths
Brian Christian is a writer with backgrounds in computer science, poetry, and philosophy, known also for 'The Most Human Human.' Tom Griffiths is a cognitive scientist and professor at Princeton (formerly UC Berkeley) who studies how human minds solve the same problems computers do. Together they translate technical computer science literature for general readers.
