Indeed. But testing in general, as it is done currently, in modern formal education, discourages deep learning as opposed to shallow learning.

Why? Because tests with marks implore students to start learning at max 3 days before the test, thus getting knowledge into short-term memory and not long term memory. Rendering the process of learning virtually useless even though they "pass" the curriculum.

I know this because I was such a student, and saw it all around me with virtually every other student I met, and I was in HAVO, a level not considered "low".

It does not help that teachers, or the system, expect students to know how to learn (efficiently) without it ever being taught to them.

My message to the system: start teaching students how to learn the moment they enter high school

Reply to Nidhi Sachdeva: Nidhi Sachdeva, PhD Just went through the micro-lesson itself. In the context of teachers using to generate instruction examples, I do not argue against that. The teacher does not have to learn the content, or so I hope.

However, I would argue that the learners themselves should try to come up with examples or analogies, etc. But this depends on the learner's learning skills, which should be taught in schools in the first place.

Reply to Nidhi Sachdeva: I agree mostly, but I would advice against using AI for this. If your brain is not doing the work (the AI is coming up with the story/analogy) it is much less effective. Dr. Sönke Ahrens already said: "He who does the effort, does the learning."

I would bet that Cognitive Load Theory also would show that there is much less optimized intrinsic cognitive load (load stemming from the building or automation of cognitive schemas) when another person, or the AI, is thinking of the analogies.

https://www.linkedin.com/feed/update/urn:li:activity:7199396764536221698/

(1:20.00-1:40.00) What he describes is the following: Most of his notes originate from the digital using hypothes.is, where he reads material online and can annotate, highlight, and tag to help future him find the material by tag or bulk digital search. He calls his hypothes.is a commonplace book that is somewhat pre-organized.

Aldrich continues by explaining that in his commonplace hypothes.is his notes are not interlinked in a Luhmannian Zettelkasten sense, but he "sucks the data" right into Obsidian where he plays around with the content and does some of that interlinking and massage it.

Then, the best of the best material, or that which he is most interested in working with, writing about, etc., converted into a more Luhmannesque type Zettelkasten where it is much more densely interlinked. He emphasizes that his Luhmann zettelkasten is mostly consisting of his own thoughts and is very well-developed, to the point where he can "take a string of 20 cards and ostensibly it's its own essay and then publish it as a blog post or article."

The essence for this video is correct; active learning, progressive summarization, deep processing, relational analytical thinking, even evaluative.

Yet, the implementation is severely lacking; marginalia, text writing, etc.

Better would be the use of mindmaps or GRINDEmaps. I personally would combine it with the Antinet of course.

I do like this guy's teaching style though 😂

(13:36) Perhaps this is also why visual information is so useful for learning and cognition (see GRINDE)... Maybe the visual medium should be used more in instruction instead of primarily auditory lectures (do take into account redundancy and other medium effects from CLT though)

When it comes to thinking, the Zettelkasten solves an important issue which is the problem of scope, which is impossible at the current moment in mindmapping software such as Concepts.

Mainly, Zettelkasten allows you gain a birds-eye holistic view of a topic, branch, or line of thought, while allowing you to at the same time also gain a microscopic view of an "atomic" idea within that thought-stream, therefore creating virtually infinite zoom-in and zoom-out capability. This is very, very, beneficial to the process of deep thinking and intellectual work.

CIFAR-10 performance results are overestimates since some of the training data is essentially in the test set.

Kozlowski, Diego, Jennifer Dusdal, Jun Pang, and Andreas Zilian. ‘Semantic and Relational Spaces in Science of Science: Deep Learning Models for Article Vectorisation’. ArXiv:2011.02887 [Physics], 5 November 2020. http://arxiv.org/abs/2011.02887.