Book: Algorithms to Live By: The Computer Science of Human Decisions by Brian Christian. Rating: ⭐️⭐️⭐️⭐️.

This is that kind of book that does not answer questions, it just creates a lot more. :)

I tend to not implement algorithms just for the sake of learning a new one and this is not a good thing. When you read these kinds of books you understand why it’s so important to know which kinds of algorithms already exists and which problems they solve.

Real hackers (Read Hackers: Heroes of the Computer Revolution) implement their own code just for the sake of learning something new or for the joy of programming. When you really need to deal with a complex problem, the tools will be already in your mind, just compose and create.

This book helps understanding some interesting algorithms that we use (or should be using) in real life. You will understand the basics about them.

In case you’re a programmer, it will probably generate a list of new topics to study. :)

Here is my raw notes about the book:

  • A mention to the secretary problem
  • Look to 37℅ before choosing, then choose the best one
  • To decide the total amount look at the total time you can spend with that search
  • A/B testing is a good explore exploit example. Look for data then explore it.
  • Base rule
  • Laplace rule
  • Gauss curve
  • It’s important to live with reliable adults. That could influence the [[marshmallow experiment]].
  • “Basyan protect your prior”
  • 9 factor model
  • Sorting is a big computing problem that we use in our daily life
  • When you just have one computer and the tasks do not depend on one another, it doesn’t matter the order you do them, they will always be finished at the same time.
  • complexity penalty
  • Problem relaxations: constraint relaxations. Traveling salesman, remove the need to not pass in the srme city moase than once
  • Continuous relaxations: transform binary or discrete into continua. When choosing between two choices use a 50-50 chance.
  • Langrange relaxation: violate the rules and add penalties.
  • Gredy algorithm? Traveler salesman
  • Hill climbing algorithm
  • A connection doesn’t exist in the internet, it’s just a concept.
  • Exponential backoff
  • AIMD algorithm (opposed to “up or out”)
  • Backchannel in a conversation
  • Being a successful investor is anticipating the antecipation of others
  • The optimist proclaims that we live in the best of all possible worlds; and the pessimist fears this is true
  • The prisoners dilemma. It’s always better to blame your partner.
  • The paradox is that if all partners blame each other, it will be bad for everyone
  • Every time you find yourself in the side of tne majority, it’s time to pause and reflect
  • Use other people’s opinions as part of your thoughts but do not use theirs only.
  • Recursion in poker. You never play your hand. It’s your hand, what you think the opponent has, what the opponent think you think it has and so on.
  • Help people to not have to compute things to be able to understand you
  • Say what you want instead of giving clues about it. To let the person simulate your needs is costly and usually doesn’t generate good results.
  • Do not delegate your problem to someone else
  • Whenever you give a choice, reduce the number of items to make it easy for the one choosing.