Algorithms to Live By

In Algorithms to Live By, authors Brian Christian and Tom Griffiths reveal how computer science offers timeless lessons for daily decision-making. Every human problem—choosing a career, managing your calendar, organizing your home, or even finding love—mirrors the computational challenges machines solve every second. Like computers, humans operate with limited memory, time, and energy, and thus can benefit from the same efficiency principles that keep digital systems running smoothly.

The book demonstrates that the logic behind algorithms can provide clarity and peace of mind in situations that once seemed chaotic. By combining mathematics, psychology, and philosophy, Christian and Griffiths show how thinking like a computer can actually make us more human —more deliberate, self-aware, and effective in an unpredictable world.

The 37% Rule: Knowing When to Commit

One of the book’s most celebrated concepts is the 37% rule, a solution to what mathematicians call the “optimal stopping problem.” Whenever you face a series of options—like hiring an employee, choosing a house, or picking a romantic partner—you must decide when to stop searching and commit.

Here’s how it works:

• Evaluate the first 37% of your options without committing.

• Once you’ve seen that sample, choose the next option that’s better than all the ones before.

Examples:

• Dating: If you expect to date between ages 18 and 40, your “exploration” phase should last until around age 26. After that, commit when someone surpasses all your previous partners.

• Hiring: If interviewing 100 applicants, reject the first 37 to set a standard, then hire the next one who beats them.

• House hunting: Tour the first 11 of 30 houses before making offers.

• Parking: On a single-lane street, skip the first 37% of available spots before choosing one.

But the rule has limits. Mathematician Hannah Fry points out that the probability of landing the absolute best option is only 37%. If “great” is good enough, you can adjust the percentage. For instance, if you’re willing to accept a top 5% choice, start committing after 22%, raising success odds to 57%.

In other words, life rarely demands perfection— satisficing, not maximizing, is often the real win.

The Explore–Exploit Dilemma: Curiosity vs. Stability

Another timeless algorithmic principle is the explore/exploit tradeoff. It answers the question: Should I keep trying new things, or stick with what works?

Computers face this every day—like streaming services recommending new shows or restaurants balancing new menu items with bestsellers. Humans do too:

• Career: Should you stay in your secure job or test a new industry?

• Relationships: Stick with a comfortable partner or meet someone new?

• Food: Try that unfamiliar café or revisit your favorite spot?

Christian and Griffiths explain that the right answer depends on where you are in life:

• Early stages: Focus on exploration. If you’re young or starting a new career, experiment widely—you’re gathering data.

• Middle stages: Split between exploring and exploiting. Try new hobbies or projects while doubling down on proven strengths.

• Later stages: Prioritize exploitation. Invest time in what already brings value, joy, and stability.

In practice, you can apply algorithms like:

• Upper Confidence Bound (UCB): Choose the option with the highest potential upside—even if its current payoff is uncertain.

• ε-greedy: Exploit 90% of the ...