How algorithms can inform our life´s decisions

“Algorithms to Live By” is a compelling read by Brian Christian and Tom Griffiths that explores what we can learn from algorithms in our decision making process. While it definitely deserves a read, here are some of the key takeaways for those too busy with their own decision making to dive into the book.

1. When to stop is one of the most important elements in decision making. To resolve the trade-off between seeking ever more complete information and making decisions with incomplete information, they introduce tools such as the 37% rule and the look-then-leap rule. Applications of these tools range from selling a house to choosing a team member or spouse to parking a car. Considering the total time affected by the choice, no more than approximately one-third of the time should be spent sampling before deciding. This can come natural when choosing a house or a job, but it may seem unnatural when choosing a partner: always sample first before making a long-term commitment.

2. When to explore and when to keep exploiting. The Gittins Index and Upper Confidence Bound are algorithms that make these decisions in machines. As humans, we can learn from them to prefer one over the other depending on the total amount of time available and our experience with the current decision. Use cases for such problems include finding new medications, big life decisions related to career and life in general, and smaller life decisions like choosing a restaurant for dinner or deciding when to keep playing a slot machine. Looking at a human life, ideally life should gradually get better, because the older you get, the better chance you have” to exploit knowledge acquired over decades”. In the same way, children behave very rationally when they explore their surroundings with all their senses at an early age so that they can later exploit what they really like.

3. When to sort and how to do so. Great news everyone: leaving a stack of documents to be processed on the table is a super efficient sort! The usefulness of sorting depends on the amount of items to be sorted, the frequency of the search, and the amount of time it takes to find a particular item.

Source: SwissCognitive

Own a fractional luxury property

Share this:

Related