7 ms·
I think the software that is promoted here is actually detrimental to "remembering". The core thesis here "Learning is intertwined with memory" was indeed hash
by fwbex 4y ago
I think the software that is promoted here is actually detrimental to "remembering".
The core thesis here "Learning is intertwined with memory" was indeed hashed out by psychologists in the 1970 (Craik and Lockhart); It was precisely these two who put forth the "Levels/depth of processing" idea, which emphasized the importance of the encoding process. A more "meaningful" encoding enhances memory.
The author not so secretly wants to sell us an app for spaced practice. And yes, spaced practice is indeed important and has been well studied in memory research. But I feel that the tool that they are selling is the antithesis of good learning, especially if you want to check "Quantum Mechanics" off of your list. If I got it right, the main difference of their app to Anki or any other flashcard program is that you do not need to spend so much time drafting your questions because an AI can do it for you. But this is precisely where deep encoding would come into play. You need to decide how to phrase that question and write it down! You need to draft that answer! The classic crib sheet phenomenon.
Regarding computerized support for such human "deep learning" I can recommend "concept mapping" tools (VUE from Tufts or ihmc CmapTools). But a pen and paper is great, too.
- p-christ 4y agoYou review the AIs suggestions which itself is a form of encoding (that is arguably more powerful than the encoding you describe)
- captaincaveman 4y agoWell shouldn't be too difficult to test these hypotheses, if the AI approach is better then actually run a study and show that!