# Asking Better Questions Asking good questions is one of the most challenging things that a [Researcher](Researcher%20MOC.md) must do. But what are "better questions"? Better questions are those whose answers can significantly change the way that we understand a phenomenon. Given millions of data points, you still have no [Knowledge](Knowledge.md) - that requires you to *extract* information from this data. This requires *creatively* asking questions that can yield new [Explanations](Explanations.md). Consider a bad question: "what fraction of bids placed in Ercot don't clear?". What would an answer of 70% instead of 80% mean? Answering this questions build us no closer to building [a true understanding, we are stuck at a descriptive level](Description%20Is%20Not%20Explanation.md). The point is not to build up a nice suite of confirmatory examples. Rather it is to build up a [generalizable](Generalization.md) understanding - an [Explanation](Explanations.md) - of the phenomenon at hand. One that can lead to insight and downstream consequences. Asking better questions is intimately tied with [Measuring the Right Thing](Measuring%20the%20Right%20Thing.md). In a very real sense, asking the right question will be one of the greatest skills one could have as we move into the future. For remember, [The World is a High Tech Oracle](The%20World%20is%20a%20High%20Tech%20Oracle.md). Tools such as LLMs will make *retrieval* ever easier (just as Google was a large improvement over needing to looking up information at a library via physical books). However, the *asking* of the right question is still inherently a [creative](Creativity%20MOC.md) task, one that as of now only humans can perform. --- Date: 20240903 Links to: Tags: References: * [In search of better questions | Patreon](https://www.patreon.com/posts/in-search-of-47047644)