Jan, 2026
You can describe things but you cannot explain them because you can never provide sufficient context needed for an explanation. The sufficiency of context depends on the person receiving and therefore it depends on him to fetch the right context himself. You are a passive participant when someone is trying to understand something and thinking you’re more than that is a form of an intellectual saviour complex. You cannot explain a leaf without having to explain photosynthesis, the tree, the air and the rest of the universe. When you even think of trying to explain something, you’ve already failed. You can try over explaining and the over explanations will ironically require even more explanations, not less. Because context is recursive. No matter how many layers of explanations you provide, each layer subtracts from an infinite number. The sufficient context is N-1 where 1 is each layer and N is all the information in the world. While you cannot explain a leaf, you can describe it. Your job therefore is to describe things as they are and as you notice them. And if a person does not understand it, they need to spend more time thinking about their own explanation. And if you’re at the end that is unable to understand something, you need to go and look for more facts and not more explanations. An example of how borrowed explanations don’t work is to try to explain a joke when someone does not get it. The explanation kills it. Most explanations are like this, they don’t work if you try to give them to others. So describe for others and explain for yourself. But you can’t create your own explanation for everything? That is true, but you don’t need to create an explanation for everything. Only for the things you value. And so the problem is actually figuring out what you value. And in a world where people don’t create explanations for anything themselves, they don’t value anything. With borrowed explanations they end up with borrowed values that contradict and end up in amusement park rides they don’t even like. A corollary for this is if you see someone create explanations for something, know that they value it. Some people explain away politics. They value politics. Some people explain away the nuances of writing. They value writing. You can find people who value the same things by looking at people who have come up with explanations for the same things. Even if a person’s own explanation is wrong, the resulting understanding is better than it would be if a more correct explanation was borrowed. Now that you know that you have to explain things to yourself, let me describe how the process works. It is childishly simple. You look at every part of a description that you don’t understand and then go look for descriptions of that part. And then if that description also has parts that you don't understand, you go do the same. From here onwards are details of this topic that are unnecessarily pedantic for you maybe but they exist because they have value to me. The problem inherent with explanations is that when a piece of information is moved from its original environment to someone, it loses context. To pass a piece of information in its entirety you have to pass the rest of the universe with it. This means that information is never complete. But we don't need information to be complete. We only need it to be sufficient. For information to be sufficient, we need access to the context in real time because the universe is too big to be preloaded with a leaf. It is not even possible to preload sufficient context for information. Because context is temporal and when it is preloaded, it has already expired when someone tries to use it to make sense of the information. The only way for information to have the right context that is still applicable is extracting context at the point when the information is being used. All of this means that our ability to get information and act on it depends not on how clean the original source is but how quickly we can extract the context from it in real time. It is much easier to see a plant and then find what it is than to memorize every plant and then try to identify the one you see because the space for possible plants is too big and you don't need the knowledge of every plant just to work with a single plant. A person who is really good at making the right decisions has the ability to extract the right context quicker than others. This is why learning from history has a problem. Ignoring the fact that it is rarely accurate, history has too much information because of the possible permutations that you might never run into a situation close enough to it to benefit from it. The present will always be different enough from the past in a way that makes the past not that useful. And this also means that there is a right level of context for every information. How do you know what level is the right level of context? You tell this based on the fact whether reality makes sense with that level of context. With an insufficient level of context, what you observe in real life does not make sense with what you understand or know. This is a hint that you need more context around the piece of information that you have and that does not make sense. All of this makes sense in parts and seems obvious because this is how basic science works. We observe parts of the world and then make sense of these observations by extracting more information from the original context. (An AI extracts context by adding more files of the codebase). We do the same in all parts of our lives. A person or a concept that does not make sense to you has a context that you don't have available to you. And oftentimes understanding a very complex idea depends just on getting that context.
If you want to be notified of new articles, this is my substack where I send out unscheduled emails containing new essays and things that I am thinking or reading.