Brinn's design is grounded in decades of peer-reviewed cognitive science. Each core product decision — frictionless capture, semantic graph, associative recall — maps directly to a finding about how human memory actually works.
Key finding
We forget approximately 58% of newly learned information within 20 minutes and roughly 70% within 24 hours. The forgetting curve is exponential — the steepest loss happens in the first hours after encoding.
Ebbinghaus's 1885 experiments on memory retention established the foundational model of biological forgetting. He demonstrated that memory loss follows a predictable exponential decay function — and that spaced review dramatically slows this decay. For knowledge workers, this means an insight captured at 9am and not revisited will be effectively lost by evening.
Key finding
Using external tools to store and process information genuinely frees working memory resources for higher-order thinking — reasoning, creativity, and decision-making.
Risko & Gilbert's 2016 review formalised cognitive offloading as a research discipline. Their central finding: when people offload storage tasks to external systems, the freed cognitive resources are reallocated to more complex processing. The act of externalising is not a crutch — it is neurologically sound. External memory tools function as a genuine extension of the cognitive system, not a replacement for it.
Key finding
Memory retrieval works via spreading activation through a semantic network — related concepts activate each other. We recall 'doctor' faster after reading 'nurse' than after reading 'bread'. Meaning drives memory, not alphabetical order.
Collins & Loftus's spreading-activation model established that human long-term memory is fundamentally associative and semantic — not hierarchical or alphabetical. When you retrieve a memory, activation spreads outward through a web of related concepts. This is why a smell can trigger a memory from twenty years ago: the sensory input activates a node in the semantic network, which activates connected nodes.
Key finding
Human working memory can hold approximately 7 (±2) items simultaneously — commonly called 'Miller's Law'. Beyond this capacity, items are dropped or confused.
Miller's 1956 paper is one of the most cited in cognitive psychology. Its core finding — that working memory has a hard capacity limit — has been refined by subsequent research (Cowan, 2001, suggests 4 ± 1 as the true limit) but the fundamental insight holds: the biological brain is a processor, not a storage device. Attempts to hold more than a handful of items in working memory simultaneously degrade performance across the board.
The takeaway
The cognitive science is unambiguous: biological memory is a processor, not a storage device. It decays predictably, has a hard capacity limit, retrieves by semantic association, and is genuinely extended — not just aided — by well-designed external memory systems. Brinn is built to be that system: frictionless enough that capture doesn't consume the cognitive bandwidth it frees, and semantically organised to mirror how the brain actually retrieves information.
Build your knowledge graph