Brian Christian, co-author of “Algorithms to Live By: The Computer Science of Human Decisions,” is on today’s podcast. The book shows how common algorithms relate to our everyday lives.
Brian starts off giving a historical explanation of an “algorithm”. Algorithms do not just relate to computers or mathematical procedures. A cooking recipe can be described as an algorithm or you could use an algorithm to solve the question, “How do you know when your best opportunity is?” Brian says you use an algorithm known as “optimal stopping” to solve this question. He explains the algorithm that provides the best probability of the best outcome.
Probability is the next. Is it worth exploring a new business possibility? Or is it better to hone in your skills and continue on with what you are doing? There is a formal framework or algorithm to help make these decisions. Brian explains the “win stay, lose shift” approach. If a restaurant is good, go back. If you don’t have a good experience, don’t go back. It may not be the optimal approach but it is a good approach and easy to live by.
Michael jumps to an example with Jeff Bezos, the CEO of Amazon. Brian tells Jeff’s story of Amazon and how he left his successful job at the time to start up this “online bookstore”. He ultimately used a “regret minimization framework” to make his decision. There was a possibility Jeff may not have been successful starting Amazon, but had he not tried it, he would have regretted not trying for the rest of his life. Brian also gives other examples of regret minimization algorithms.
Michael and Brian discuss threading next. Lastly, Michael brings up the “searching vs. sorting conundrum” that Brian discusses thoroughly in his book. Brian gives examples about sorting through information on the computer. He says that people should ask themselves “Should you be sorting at all?” Brian explains why “messy” is sometimes better.
In this episode of Trend Following Radio:
- Optimal stopping
- What is an algorithm
- Process vs. Outcome
- The explore, exploit trade off
- The multi-arm bandit problem
- Win stay, lose shift
- Regret minimization framework
- Frequency and intensity of mistakes related to age
- Upper confidence bound algorithm
“There is this argument that could be made, in a number of situations, where mess is not only the easy choice but it can be the optimal choice.” – Brian Christian