- Topics: [[Problem-solving]] - People: [[Tom Griffiths]] --- ## Summary - The [37% rule](https://en.wikipedia.org/wiki/Secretary_problem), the [explore-exploit trade off](https://en.wikipedia.org/wiki/Multi-armed_bandit), and the principle of [least recently used](https://en.wikipedia.org/wiki/Cache_replacement_policies). - **Do the best you can** do by sometimes taking a chance and by sometimes failing, but **always learning**. That's what being rational means. - [[You can't control outcomes, just processes]]. - [[It is not a problem to solve. It is a process to manage. | Life is not a problem to solve. It's a process to manage.]] - Taking a chance means learning something new - about yourself and the world you live in. It is gathering information that you can later use. - [[The value of information increases the more opportunities you have to use it]]. - Doing what's optimal may seem "boring". That's because all curious and potential paths have already been explored. The only path remaining is the optimal **desired path**. ## Notes ### 37% rule - How do you switch from inaction to action? The [37% rule](https://en.wikipedia.org/wiki/Secretary_problem). - "Trying to find a place to live is an [optimal stopping problem](https://en.wikipedia.org/wiki/Optimal_stopping)." ### Living like a computer - "...think about the computational structure of problems that arise in everyday life, and compare the ideal solutions to those problems to the **way we actually behave**." - As much as we may want to computationally optimize with rational thinking, there are just sometimes that cannot be scientifically or mathematically quantified and optimized for. **Some things that should be experienced emotionally rather than considered cerebrally**. - "Living like a computer doesn't sound like a lot of fun." ### Explore-exploit trade off - To explore and try something new, gather new information. Or, doing something familiar, exploiting the information that you already have. - Examples: Trying a new restaurant. Making new friends. Buying something new to wear. - When deciding what restaurant to go to, "first ask yourself, how much longer you're going to be in town." - Short time -> Exploit. - Long time -> Explore. - [[The value of information increases the more opportunities you have to use it]]. - "A baby will stick anything in their mouths... They're exploring... The old guy, who always goes to the same restaurant and always eats the same thing **isn't boring** - **he's optimal**". - The boringness of being optimal resides in it's predictability. There are no more exploratory choices to make as those curiosities have already been answered. **The familiar and desired path is an optimal one**. - "You don't have to go to the best restaurant every night. **Take a chance**. Try something new. **The information you gain is going to be worth more than one pretty good dinner**." ### Principle of least recently used - In computers, RAM (fast memory), has to decide what existing information it has to remove in order to load new information. This is due to it's limited capacity - not unlike a closet or wardrobe. - Some strategies include: Randomness, "first in first out", and **least recently used**. - "Maybe it's worth applying the least recently used principle to organize your wardrobe as well." - Remove the thing that you haven't used (accessed) for a (really) long time. Chances are, you've replaced it with something better, or you never really needed it in the first place. - [[Make the things you're most likely to need the most accessible]]. - "Typically maligned as messy and disorganized, a **pile of papers** is, in fact, **perfectly organized**." As long as their sorted using the least recently used principle. - In isolation, these piles of papers, these **bundles of chaos**, are perfectly workable and are orderly enough. It becomes **unmanageable** when you start forgetting what goes where, often because there are **too many** bundles of chaos. This is also the case if someone else unfamiliar with your system needs to work with your stuff. In either case, failure and frustration, that dreaded **feeling of lost**, arises when context is missing or forgotten, and there is no fallback system available to support it. - "The best algorithms are about doing what make the most sense in the least amount of time." ### Simplification - "Computers face hard problems... by making them into simpler problems. Solving simpler problems... can give you insight into harder problems and sometimes produce pretty good solutions in their own right." ### Take a chance - (37%) "You fail most of the time. But that's the best that you can do." - [[You can't control outcomes, just processes]]. - "As long as you've used the best process, you've done the best that you can ." - Taking a chance and not considering all options "aren't the concessions that we make when we can't be rational - **it's what being rational means**." ## Original ![[3 ways to make better decisions - by thinking like a computer-1.jpg]] ![[3 ways to make better decisions - by thinking like a computer-2.jpg]] ![[3 ways to make better decisions - by thinking like a computer-3.jpg]] ![[3 ways to make better decisions - by thinking like a computer-4.jpg]]