Use it or lose it — thinking skills in the AI era.

The skill-decay literature distinguishes between cognitive skills (procedural and declarative knowledge needed for thinking tasks) and motor skills (procedures requiring physical execution). Both decay without practice, but at different rates. Arthur et al.'s 1998 meta-analysis found cognitive skills tended to decay faster than continuous motor skills over equivalent retention intervals — a result with practical implications for any field (aviation, medicine, military training) where infrequent-use cognitive procedures need to be reliably available years after initial training.

The use-it-or-lose-it framing is widely used in skill-acquisition research and in everyday coaching across athletics, music, language, and writing. The common pattern is that skills decay when not practised and recover when practised again — usually faster the second time, a phenomenon called the "savings effect" (Ebbinghaus, 1885, formalised). The recent research line on cognitive offloading extends the same intuition to thinking acts that are easy to outsource — recall (Sparrow, 2011), spatial memory (Dahmani & Bohbot, 2020), basic calculation, drafting, and reasoning.

More recent work on AI dialogue systems specifically asks what happens to students' cognitive practice when they lean heavily on AI as a substitute for working through problems themselves (Springer 2024 overreliance paper; EDUCAUSE 2025). The pattern is consistent with the older skill-decay framing: when the tool does the work, the underlying skill gets fewer reps, and the skill atrophies on the timeline the practice-frequency would predict.

On daily practice specifically, the deliberate-practice tradition (Ericsson, Krampe & Tesch-Römer, 1993; the partial replication by MacNamara & Maitra, 2019) supports a narrower claim than popular versions imply: deliberate practice maintains and grows the specific skill being practised, with domain specificity and diminishing returns. The implication for use-it-or-lose-it is direct: short, regular, deliberate practice is the most economical way to keep a skill in active use.

The principle is not a medical recommendation. It does not claim that practising in an app prevents dementia, slows aging, or improves clinical outcomes. Senwitt does not make any of those claims. The point is much smaller: regular practice keeps the practice habit alive.

The skill-decay literature does not say that lost skills are gone forever. The savings effect is real — relearning is typically faster than original learning. The literature also does not say that all skills decay at the same rate, or that any one app or intervention prevents decay across categories. Decay is a property of practice frequency, not of any specific product.

For an individual using AI tools every day, the practical question is whether the act of writing the sentence, doing the calculation, predicting the code behaviour, recalling the fact, reading attentively, and reasoning through a small problem still happens regularly. If it does not, the practice is on the way out. Senwitt is built to keep that practice on the calendar.

The practical guidance is small: pick three to six Skills, do one daily Set, keep doing it. The frequency is more important than the depth. Seven minutes a day is enough to keep the skills in active use; missing a day does not reset months of practice; missing a year does not require starting from scratch (the savings effect ensures relearning is faster than original learning). The product's shape — short, daily, bounded, forgiving — is tuned to what the skill-retention literature actually supports.