VI
Mark học AIGiải mã siêu trí nhớ · Mai Tường Vân · written in Vietnamese, no English edition
A super memory is not magic. It is a set of techniques anyone can learn: turn information into images, link the images together, then place them somewhere familiar.
Three chapters, with most of the weight in chapters 2 and 3. Proportions are estimated from the table of contents.
The four techniques stack: the palace only works well once you have images (1), images that "talk" to each other (2) and a ready code table (3).
| Challenge in the book | Techniques combined | Everyday use |
|---|---|---|
| 52-card deck | coding + palace | PINs, account numbers |
| Binary digits | turn 3 bits into 1 digit, then code | long numbers |
| 100 QR codes and 100 faces | imagination + association | names of new people |
| Abstract images, real photos | find an odd detail, give it an image | rarely |
| Vocabulary, essays | image keywords + palace | exams, languages |
The author says that after about three months of training she memorised 1,000 digits in 58 minutes and a deck of cards in about 2 minutes.
The book's main claims checked against memory research. Tap a row for the reason.
Dresler et al. (Neuron, 2017): novices trained 30 minutes a day for 6 weeks; words recalled rose from about 26 to 62 out of 72, and most of the gain remained after 4 months.
Maguire et al. (2003) scanned 10 superior memorisers: normal intelligence and brain structure. The difference was a spatial strategy; 9 of 10 used the memory palace.
Meta-analysis of over 300 experiments (Cepeda et al., 2006). Dunlosky et al. (2013) rate spaced practice and self-testing as "high utility".
Richards and Frankland (Neuron, 2017): forgetting drops needless detail and helps generalisation. Ebbinghaus's forgetting curve was replicated in 2015.
Place cells and grid cells in and around the hippocampus, Nobel Prize in Medicine 2014 (O'Keefe, May-Britt and Edvard Moser).
The effect is real when odd images are mixed with ordinary ones, but fades when everything is odd (McDaniel and Einstein, 1986). Interaction between images is the more reliable factor.
Strong right after learning (Atkinson and Raugh, 1975), but some studies find faster forgetting than rote after a week without review (Wang and Thomas, 1995). Dunlosky (2013) rates it "low utility" for learning in general.
That is the author's own result, with a coach. Research shows everyone improves, but at very different speeds. Preparing for a world championship usually takes hours a day for months.
According to one review, that is the book's definition. Psychology counts short-term memory as seconds up to about 30 seconds, holding roughly 4 to 7 items. Anything still there after weeks is long-term memory.
The book doesn't promise this, but readers easily assume it. The skill is narrow: a man trained to recall 80 digits (Ericsson and Chase, 1980) still recalled only about 6 letters. Memory athletes lose their keys like everyone else.
Read chapter 2 and do the exercises. Use the techniques for things you must recall word for word, not as a substitute for understanding.
One reviewer compared it with Joshua Foer's Moonwalking with Einstein and found it thinner on science, simplified to the point of inaccuracy in places. It also skips episodic and procedural memory, both directly relevant to the memory palace and to learning.
Joshua Foer, 2006 US Memory Champion, wrote that after the contest he hardly used the techniques, since paper and phones exist. They pay off in narrow places: vocabulary, formulas, speeches, names.
For students, the biggest gain probably comes from what the book says less about: self-testing and spaced review. Both work for every subject, including the parts you need to understand.
One round of learning 20 new foreign words, combining all four techniques and a review schedule.