There are 3 major theories behind how we have designed and will evolve RECIPEs as we learn more about it: Meta-cognition (learning about learning), Contextual Learning (generalizability of learning through appreciating concepts surrounding the main concept), and Experiential Learning (learning through doing).
This is the theory of how we are able to think about our thinking, and therefore learn about our learning, and strategize it. As humans we possess this capability to reflect on how we learn some things more easily, and some ways of learning work better for us etc. This helps us plan how we learn various topics. A conscious effort and framework to develop this "muscle" further is a very worthwhile investment because it supercharges our ability to learn how to learn, rather than just learning stuff outside of their context which has the issue of lack of generalizability.
Meta-cognition has 3 components:
This is the type of knowledge that can be declared in statements, aka facts. Concepts and assets we document on SNAPlists are good examples of this.
This is the "know how" knowledge, aka, steps of carrying out a task. The interesting thing about this type of knowledge is that it is harder to declare. The canonical example is that it is very hard to explain to someone how to ride a bike, and it's much easier to show them and have them practice themselves. On SNAPlists, the fact that we put resources in order, and provide some notes about how to use them is there to document how you learned the topic and think others should too. Think of it as showing others how to learn that topic. Another important aspect of this is where we document potential use cases because without that context the list is less meaningful.
This is when you use a combination of content and procedural knowledge to plan for a particular learning objective. The overall narrative of your list should be strategic in the sense that everything - concepts, resources, annotations, order - should fit into each other to create a coherent story about how one could achieve the learning objective of the list. Ultimately, you are telling a conditional story, "if you want to learn this, this is how you can do it!". There are more complex layers of conditioning, but it's coming up in the future designs and releases.
Learning within context is far more impactful than learning things independent of what else needs to be learned around them. It both helps the learner get what is happening better, but also to retain that information for longer through association. The little concept graph we build in SNAPLists is an important part of this by showing the background concepts relevant to the main concept in an immediately visual way.
Ideal SNAPLists enable the user to learn by doing. In other words, it enables them to do something rather than just providing FYI type of info. For example, they provide a step by step guide on How to Tune Your GAN Model, or How to build a ChatBot, or How to use bias frameworks in the context of natural language processing.