Audiobook Summary and Review by StoryShots
Following the optimal math strategy still means failing sixty-three percent of the time.
Most advice tells you to weigh every option carefully before deciding.
Computer science proves that is often the worst thing you can do.
That is the argument behind Algorithms to Live By: The Computer Science of Human Decisions, by Brian Christian, and it turns apartment hunting, dating, and scheduling into solvable math problems.
Most people believe good decisions come from gathering every possible option before choosing.
Picture apartment hunting: you tour twenty units, terrified of settling too soon or missing something better down the road.
Computer scientists solved this exact problem decades ago.
Spend the first thirty-seven percent of your search gathering data, rejecting everyone, no matter how good.
Then commit to the next candidate who beats everyone you have already seen.
This is not a hunch.
It is math.
Thirty-seven percent of your search should be pure looking.
The rest should be pure leaping.
The 37% rule only tells you when to stop looking within a single decision.
It says nothing about how to spend the thousands of small choices that make up an entire life.
Move to a new city for one year and your first dinner out is guaranteed to be the best restaurant you have visited so far, simply because it is the only one you have tried.
Returning to it forever means never finding something better.
Trying somewhere new every night means never enjoying what you already know is good.
Computer scientists call this the explore-exploit tradeoff, and it governs everything from online ads to clinical drug trials to why older people call the same three restaurants while twenty-year-olds try something new every week.
Exploring costs you certainty.
Exploiting costs you discovery.
Neither strategy works alone, and standard advice, just balance them, tells you nothing about how.
The unknown has a chance of being better even when you expect it to be no different at all.
Something called the Gittins index determines exactly when exploring stops paying off, and it reshapes how you should treat your remaining years.
Picture two slot machines.
You do not know which pays out more, so every pull teaches you something while costing you a possible better payout elsewhere.
Mathematicians proved that the optimal strategy is not to track expected value at all.
It is to always play the machine with the highest possible upper bound, the one whose uncertainty gives it the best chance of being secretly great.
This resolves the restaurant dilemma precisely: explore aggressively when your time horizon is long, because unknown options carry hidden upside.
Exploit relentlessly when time is short, because you have already gathered what you need.
Novelty-seeking fades with age not from timidity, but from a shrinking runway.
Following the optimal strategy in these problems still means failing most of the time.
The math simply guarantees you fail less than everyone guessing blindly.
If this changed how you think about decision-making under uncertainty, someone in your life navigating a big choice right now would probably want this too.
This summary of Algorithms to Live By threads together the 37% rule, the explore-exploit tradeoff, and the Gittins index into one argument: your hardest life decisions already have mathematically proven answers hiding in computer science.
The full summary covers sorting theory and why your messy desk might already be optimal, caching theory and what your closet has in common with your computer's memory, and the surprising role of randomness in escaping bad decisions entirely.
You will also get the scheduling math behind procrastination and why answering emails immediately might be sabotaging your actual output.
Anyone drowning in choices, deadlines, or indecision needs this.
For the full summary of Algorithms to Live By by Brian Christian, plus the infographic and animated video breakdown, head to the StoryShots app.