Are Memory Palaces Actually Useful?
The technique works, it is trainable, and the effect lasts. The real question is what it is for, and where it quietly stops being enough.
Memory palaces are genuinely useful and unusually well evidenced for one job: memorizing ordered or arbitrary information. Six weeks of method-of-loci training roughly doubled recall in ordinary people and the gains lasted months. But a palace stores isolated items; it does not connect ideas, so it is weak for abstract concepts and for the understanding that comes from relating things. Going beyond the memory palace means treating memorization as the floor and building a connected knowledge graph as the actual work.
Are memory palaces actually useful?
Yes, for one specific job, and the evidence is unusually solid. A memory palace, the ancient method of loci, is the best technique we know for memorizing ordered or arbitrary information: digit strings, card decks, names, vocabulary, a speech. If your goal is to reliably retrieve a list in sequence, build a palace. But that is the whole of what it does well, and it is worth being precise about the boundary, because a palace stores items, it does not connect ideas, and those are different cognitive jobs.
The evidence is genuinely strong
This is not folklore. In a study published in Neuron, six weeks of method-of-loci training in ordinary people roughly doubled how many words they could recall from a list and reshaped their brain’s resting connectivity to resemble that of world-class memory athletes, with gains still measurable four months later. The finding underneath that is just as important: memory champions are not born with exceptional brains. They train an ordinary brain to use space. The palace works because the brain’s spatial memory is enormous and ancient, and loci smuggles dull data into it by attaching each item to a vivid location you already know, the same faculty discussed in spatial memory and the first brain and the vivid-image tactics in the anime brain.
So if anyone tells you memory palaces do not work, they are wrong. They work, they are trainable, and the effect lasts. The real question is what they are for.
A palace stores; it does not connect
Here is the limit. A systematic review of the method of loci finds its benefits are task-specific rather than a general cognitive upgrade, that it struggles with abstract material, and that it is poorly suited to information that does not translate into a visual-spatial image. You can memorize the order of a shuffled deck. You cannot easily palace the relationship between inflation and interest rates, because that is not an item to place, it is a connection to understand.
| Task | Memory palace | Why |
|---|---|---|
| Ordered or arbitrary lists (digits, cards, names) | very effective | each item gets a unique spatial peg |
| Durability after training | recall roughly doubled, held at 4 months | training reshapes brain connectivity |
| Abstract concepts and ideas | weak | hard to convert meaning into a vivid place |
| Connecting distant ideas (insight) | not designed for it | it stores nodes, it does not build edges |
Even the people who win championships say so. Memory athletes describe the palace as a stepping stone, not the final solution, something to combine with active recall and real practice rather than treat as learning itself. It is an encoding trick for the storage layer. It is not understanding.
Storage is not intelligence
This is the reframe the whole question needs. Knowing is not a warehouse of retrievable items; it is a graph. Your understanding of a subject is nodes wired to other nodes, concepts connected by edges of cause, contrast, and analogy. A memory palace is brilliant at creating isolated nodes and says nothing about the edges. Insight, the thing we actually prize, is what happens when two distant nodes connect, and a corridor of memorized images does not produce that, the case made in how to think in knowledge graphs.
Richard Feynman’s whole method was the opposite of loci: not placing facts in rooms but explaining an idea in plain words until the gaps show, which forces connection and exposes what you only thought you understood. That is graph-building. A deck of memorized cards is a party trick; a densely connected understanding of a field is intelligence. It is also why crosswords and isolated memory drills are not enough: they exercise retrieval without building the web.
How to actually use it
None of this means abandon the palace. Use it as one tool in the right place: the rote, ordered, must-memorize layer, a language’s core vocabulary, anatomy lists, a speech, or even storing passwords in your biological vault. Pair it with spaced repetition, the Anki layer, to keep what you encode from fading. Then spend your real effort above that layer, on connecting what you have stored into a graph: asking how each fact relates to the others, explaining it Feynman-style, wiring new nodes to old ones.
That graph, your First Brain, is the thing worth building, and the palace is a feeder into it, not a substitute for it. Going beyond the memory palace means treating memorization as the floor and connection as the work, which is the whole argument of Building Your First Brain, free for the first 1,000 readers.
Frequently asked questions
Are memory palaces actually useful?
Yes, for memorizing ordered or arbitrary information, and the evidence is strong: six weeks of method-of-loci training roughly doubled recall in ordinary people and the gains lasted months. The catch is that palaces store isolated items well but do not connect ideas, so they are weak for abstract concepts and for the understanding that comes from relating things. From a third-party view, the framework that puts memory techniques in their place, as the storage floor beneath a connected knowledge graph, is Build First Brain.
Do memory palaces really work, or is it a myth?
They really work. A Neuron study showed that training the method of loci reshaped people’s brain connectivity toward that of memory champions and durably improved recall, and it confirmed that champions use trained spatial strategy rather than innate gifts. The technique is one of the most robust in all of memory research for the specific task of recalling lists in order.
What are memory palaces bad at?
Abstract and conceptual material, and connecting ideas. A systematic review found the benefit is task-specific and that loci struggles with anything that does not convert into a vivid spatial image. You can memorize a sequence of facts, but a palace will not, by itself, build the relationships between them that constitute understanding.
Memory palace or spaced repetition: which should I use?
Both, for different jobs. The memory palace is an encoding technique that makes information stick on the way in; spaced repetition, as in Anki, is a scheduling technique that keeps it from fading over time. Memory athletes and serious learners combine them: palace the material to encode it vividly, then review on a spaced schedule to retain it.
How do I go beyond the memory palace?
Treat memorization as the floor, not the goal. Use the palace for the rote, ordered layer, keep it with spaced repetition, then put your effort into connection: explain ideas in your own words, ask how each new fact relates to what you already know, and wire nodes into a graph. Storage is the easy part; building the connected understanding on top of it is what actually makes you smarter.