What was the last thing you forgot? A name, a place, maybe your keys? How did you ultimately remember that information?
Recall can be a frustrating process. You know it’s there… it’s right on the tip of your tongue. But the harder you try, the more elusive memories seem to become.
Often what happens is you give up. The conversation moves on. But, in the background, your subconscious chews on the problem for a while, spitting up the answer a few minutes later.
The two stage theory explains this behavior. First, your brain searches working memory (equivalent to RAM). If it draws short, your brain begins a background associative search, sifting through cold-storage until it finds the answer. This process is slow, lossy, and error prone - not ideal.
Does your brain use hierarchy to remember things? Sort of, but not in the strict way filesystems use. It comes back to associations - we group memories into areas, not just in the categorical sense, but in the very real physical sense. This explains why, when you switch between rooms, you can sometimes resume a trail of thought.
So if we want to design tools to supplement our memory we should extend this pre-existing architecture: limited hierarchy and primarily association-based.
That’s how we designed Reflect.
In Reflect we have notes, and associations between those notes (we call these backlinks). Notes model your brain’s neurons, and backlinks model synapses. Simply type ‘[[‘ to create a backlink between notes.
For example, you might have a note related to the concept of Sequoia, the venture-capital firm. Note the purple ‘Venture Firm’ backlink, linking from Sequoia back to the Venture Firm note.
As soon as you backlink Venture Firm, a note will automatically be created. As you work with more venture firms, tagging them with the same backlink, a network of associations will be built up between those specific firms and the general concept of a venture firm.
As you add more to this network, you might create the concept of a Venture Capitalist, associated with a particular firm. For example, here we have a person (Matt N), associated with a firm (Spectrum), and an entity (Venture Capitalist).
Over time these notes and associations build up and up, forming a neural network that mirrors the inside your brain. Now, you have a super-power: Instant infallible recall.
For example, you could instantly find a list of every VC firm, and who you interacted with. Or quickly look up how you know a particular person, and which companies they’re associated with.
Note how this structure complements how your memory works. If you can’t remember a partner’s name, but you can remember the firm’s name - that’s enough. Or if you can’t remember the firm or the partner’s name, but you remember the firm is a Venture Firm, that’s enough. You can trace the associations, beginning with what you can instantly recall, until you find what you’re looking for.
Imagine what would be possible if your recall was a magnitude better than the average human’s.