The Google Effect and digital amnesia.

The 2011 Sparrow et al. Science paper ran four experiments examining recall when participants believed information would be saved (and retrievable) versus erased. The headline finding: when participants expected future access to information, recall of the information itself was lower, while recall of where it was saved was higher.

The 2024 Gong and Yang meta-analysis in Frontiers in Public Health synthesized findings across 35 subsequent studies and confirmed the effect is robust across populations and task types. The differential allocation between "what" and "where" is consistently observed.

The research does not show that search-engine users lose general memory ability. The brain treats reliable external storage as part of a distributed memory system — the same logic that applies to transactive memory in long-term couples (you don't remember everything because your partner remembers some of it).

It also does not show that the effect is pathological. "Digital amnesia" is a metaphor, not a clinical diagnosis. Real amnesia is a specific neurological condition; the Google effect is an adaptive allocation pattern.

For an AI-era reader, the Google effect is the cleanest existing research precedent for what cognitive offloading to LLM assistants looks like. AI assistants extend the same allocation logic from facts (which search engines handle) to drafting, reasoning chains, and generated outputs.

The practical conclusion is consistent: external storage is fine, even useful — the question is which skills you want to keep practicing anyway. That practice keeps the skill in regular use. Senwitt's daily Set is the practice mechanism for the six Skills we think most adults want to keep available.