TIL about the search inference framework. Search what? Like search stuff online?

Nope, its a fancy way to think about thinking that we do daily. When we are thinking about something, there are 3 things we are searching for :

  • Possibilities
  • Goals
  • Evidence

But before we dig deeper into these, lets think about what thinking really is. What do you do when you think? Some of it is just a mixture of random thoughts that appear in the consciousness our of nowhere. But a lot of thinking is directional. You start with a doubt or a question that you want an answer to. You start searching for an answer.

Let’s imagine you are looking for something to watch on the internet while eating dinner. There are multiple possibilities here:

  1. You can watch Last Week Tonight that your friend recommended recently. This possibility came from outside influence.
  2. You remember liking NerdWriter videos you have enjoyed before. This is an inside possibility came from inside, your memory.

How do you evaluate these possibilities? This is where Goals come in. Possible goals are:

  1. You wanna watch something funny.
  2. You want to watch something that doesn’t require lot of focusing on the video so that you can instead focus primarily on food.
  3. It should be something that finishes in ~ 15 minutes (average time you take to finish dinner).

Considering these goals, you also need Evidence that helps you choose the best Possibility. Some potential list of evidence could be:

  1. Friend said that Last Week Tonight is really funny.
  2. Your memory tells you that Nerdwriter videos are involved and sometimes require proper attention.
  3. Visually, you also notice that latest Last Week Tonight video on YouTube finishes in almost the time it takes to finish the dinner.

But other than these three, we also use our judgment in deciding what evidence qualifies as a “good” evidence. The above example is simple to analyze, but in complex scenarios related to let’s say, spending lot of money to buy something, we may give higher or lesser weight to a particular evidence for a particular possibility. This weight assignment is a function of our judgement.

But what good is any inference or judgement if we don’t use it to draw an inference? We devise a method to choose among multiple possibilities (action, beliefs or personal objectives) and we look for “good” evidence and then we use that evidence to choose a particular possibility. That’s how we try to reach our goals.

Think about your thinking about different things (meta thinking) using this framework. It really deconstructs the process that we take for granted and eventually helps in underscoring different components that form our thought process.