01
Stop at the right time
Optimal stopping says the perfect search can be worse than a timed search. Explore roughly the first 37%, then choose the next option that clears your best benchmark.
Brian Christian and Tom Griffiths
Algorithms
to Live By
A field guide for translating computer science into human judgment: when to stop searching, when to explore, when to sort, and when to accept that the best answer is good enough.
Decision Specimen
Human Runtime
Life is full of hard calls because time, attention, memory, and certainty are finite.
37%
Explore first
Sample before you commit
N log N
Sort only when it pays
Organization has a cost
LRU
Cache what recurs
Memory favors the recently useful
SAT
Satisfice under pressure
Good enough can be optimal
Bug
Searching forever feels responsible
Patch
Choose the stopping rule first
The Premise
Computer science is not just about machines. It is a vocabulary for tradeoffs.
Algorithms to Live By treats everyday life as a series of computational problems. Finding an apartment, deciding when to settle down, organizing a desk, planning a day, and leaving options open all carry hidden costs.
The relief of the book is not that life becomes mechanical. It is that some anxiety is just a missing rule. Once you know the shape of the problem, you can stop pretending unlimited search, perfect certainty, or total order are free.
02
Explore before exploiting
Early information has compound value. Later, repetition has value. The art is knowing when curiosity should give way to commitment.
03
Make computation kind
Sorting, scheduling, and memory all have costs. A messy desk, a short queue, or a good-enough answer may be smarter than perfect control.
Interactive Desk
Convert a messy choice into an algorithmic memo.
Pick a life problem, then tune the horizon, uncertainty, switching cost, and current option quality. The desk turns the book's core algorithms into a practical stopping rule.
Explore budget
14
Acceptance bar
74
Selected algorithm
Optimal stopping
Decision context
Algorithmic memo
Sample first, then commit.
Use the first 37% of the search to set a benchmark, then take the next candidate that beats it.
Explore
Threshold
Pressure
Desk note
Treat the early search as data, not failure to decide.
Framework Anatomy
Five algorithms hiding in plain sight.
The book's world is less about coding and more about choosing the right amount of effort for the problem in front of you.
01
Optimal stopping
Search has a price. Benchmark first, then commit when a strong option appears.
02
Explore/exploit
Novelty is valuable early. Later, harvest what already works.
03
Sorting
Order matters when retrieval repeats. Otherwise, sorting can be wasted theater.
04
Caching
Keep the recently useful close. Forgetting can be an efficiency feature.
05
Scheduling
Do the short urgent job first when queues are painful; protect deep work when context switching is expensive.
Reader Marginalia
Community Insights
"The right time to stop searching is usually earlier than your anxiety wants and later than your impatience wants."
resonated with this
"Explore when information is still changing your model. Exploit when novelty is only delaying courage."
resonated with this
"Sorting is not free. Sometimes a little mess is the price of moving quickly."
resonated with this
"Memory works better when it keeps the recently useful close and lets stale things drift away."
resonated with this
"Satisficing is not settling. It is respecting the cost of computation."
resonated with this
"Overfitting happens in life whenever you design a rule around yesterday's noise."
resonated with this
"A calendar is a scheduling algorithm wearing social clothes."
resonated with this
"Computational kindness means making the next person's problem easier to solve."
resonated with this
Practice File
Action Steps
01
Write a Stopping Rule First
Before your next search, define the sample size, the minimum acceptable bar, and the point where you will stop comparing. Do this before seeing more options.
I'll do this
02
Run One Explore Batch
Pick an uncertain area and gather a small batch of new signals: five apartments, three career conversations, or ten customer interviews. Then decide what changed.
I'll do this
03
Stop Sorting Low-Value Clutter
Choose one area where organization has become procrastination. Sort only the items you retrieve weekly and leave the rest in a simple catch-all system.
I'll do this
04
Build a Personal Cache
Move the tools, notes, people, and decisions you reuse most into the easiest access layer. Let rarely used material require an extra step.
I'll do this
05
Lower the Proof Burden for Reversible Choices
Mark a decision as reversible or irreversible. If it is reversible, set a good-enough threshold and choose faster than your perfectionism prefers.
I'll do this
06
Audit Your Queue
List today's tasks by duration and context switching cost. Batch the similar ones, clear a few short jobs, then protect one long uninterrupted block.
I'll do this
Closing Note
"A good algorithm does not make life mechanical. It gives judgment a sharper edge and anxiety a smaller job."
HourLife distillation
Back to LibraryTake it with you
Downloads & Shareables
Print it, pin it, post it. Ways to take Algorithms to Live By off the screen and into the world.
Printable · PDF
Action Checklist
Every action from this page as a printable to-do list with a 7-day tracker.
Download PDF →
Social · Image
Book Summary Card
Shareable 1200×630 card with the book and its top-voted insight. Perfect for social.
All Sizes · Gallery
Resource library
Preview and download the summary card plus every quote card in 6 sizes — Instagram feed, Story, Pinterest, YouTube thumbnail, phone wallpaper, and OG share.
Browse
→