
200,000 people test Apple's health AI before it ships
Apple will recruit 200,000 people to study health AI trust before shipping anything — while 300M people a week already ask ChatGPT health questions.
Tag archive

Apple will recruit 200,000 people to study health AI trust before shipping anything — while 300M people a week already ask ChatGPT health questions.

Common BMI mistakes and limits: why athletes, older adults, pregnancy and children need care, and how cm/kg and ft/lb mix-ups ruin a result.
A running plan is just a pacing spreadsheet with opinions. Here's how I generate one from a race date and real mileage instead of vibes.
Sleep tracking apps love a score. I wanted to know which nights actually moved the trend, so I built a logger that shows the week, not the verdict.
The same inverted word used twice with opposite verdicts. The turtle that does not eat.
Stillness is not in the character; a mountain put it there. Stopping at the back, the place you cannot see.
含 is to hold in the mouth; 握 is to clamp the thumb inside. Why the infant's strength comes from not having developed.
攝 is a hand pulling in; 地 is the thick sunken layer. Why over-provisioning is a way into 死地.
I counted the Class I entries in the 600 most recent FDA drug recall records. There are 33 of them,...
Most wellness advice arrives as a list of fixes: do this, avoid that, measure everything. Yang sheng,...
A calm habit tracker is mostly a check-off box and a streak counter. The weekly review is the feature that turns checking into noticing.

A controlled Lancet Digital Health trial of 12 consumer AI models found labeled pharmaceutical ads lifted selection of the promoted drug to 47.6 percent from 35 percent, while models rarely said an ad was present.