Closed Loop Alertness
I would love to study alertness and optimize people's time spent alert (maybe even revolutionizing how we schedule workdays for optimal individual productivity, and/or measure chronic fatigue/long covid in partnership with psychology).
To do this requires (1) measures of alertness from physiological sensors (i.e. pupil and blink tracking; see the wearable hardware projects), but also (2) a waterbottle that can measure and/or dose you specifically with caffeine and l-theanine. With this closed loop system in place we can track caffeine intake, alertness, blood glucose levels, and sleep, and start to build sophisticated predictive models of alertness where we control one dimension of the inputs (caffeine intake). Caffeine is obviously interesting as we habituate to it. This logic could work for all kind of drugs which are meant to have psychoactive effects, especially over longer periods.