break ambitious goals into verifiable steps
大多数人认为AI擅长处理整体目标和复杂任务,但作者暗示即使对于宏大的目标,也应该将其分解为可验证的步骤。这与当前AI应用中常见的'一次性解决复杂问题'的思路相悖,暗示长期项目需要更结构化的方法。
break ambitious goals into verifiable steps
大多数人认为AI擅长处理整体目标和复杂任务,但作者暗示即使对于宏大的目标,也应该将其分解为可验证的步骤。这与当前AI应用中常见的'一次性解决复杂问题'的思路相悖,暗示长期项目需要更结构化的方法。
Good video. Funnily enough, I related it to Mazlow's hierarchy of competence a minute before you mentioned it. (Mr. Hoorn here, btw.) Another connection I made was to van Merriënboer et al. their "Ten Steps to Complex Learning" or "4 Component Instructional Design". Particularly with regards to doing a skill decomposition (by analyzing experts, the theory, etc.) in order to build a map for how best to learn a complex skill, reducing complexity as much as possible while still remaining true to the authentic learning task; i.e., don't learn certain skills in isolation (drill) unless the easiest version of a task still causes cognitive overload. Because if you learn in isolation too much, your brain misses on the nuances of application in harmony (element interactivity). Related to the concept of "the whole is greater than the sum of its parts". You can master each skill composite individually but still fail epically at combining them into one activity, which is often required.
( ~ 5:00 )
The first stage of learning a complex skill is creating relevance, not in the sense of making knowledge relevant to your life; but rather in seeing what is relevant to learn at this point in the learning career.
Building a map...
The actions are exploration and challenge. Exploration = getting diverse opinions from others and learning the theory & variables. Challenge = open-mindedness for other beliefs and assumptions.
Reminds me of 10 Steps to Complex Learning for curriculum design, where doing a skill decomposition is one of the first steps in designing the curriculum, and either being an expert or having access to experts is paramount.
the complexity of the problem will defeat us unless we find a simple way of writing it down, which lets us break it into smaller problems.
Templates are prone to unnoticed runtime errors, are hard to test, and are not easy to restructure or decompose.
In contrast, Javascript-made templates can be organised into components with nicely decomposed and DRY code that is more reusable and testable.
Dudel, C., Riffe, T., Acosta, E., van Raalte, A. A., Strozza, C., & Myrskylä, M. (2020). Monitoring trends and differences in COVID-19 case fatality rates using decomposition methods: Contributions of age structure and age-specific fatality [Preprint]. SocArXiv. https://doi.org/10.31235/osf.io/j4a3d
Decomposing Affine Transforms