Your smartphone can actually tell if you’re stoned. No, you’re not being paranoid

Your smartphone can predict if you're high with alarmingly high accuracy

No, you're not being paranoid - your smartphone can actually tell if you're stoned. That's the findings of a new study with researchers discovering that the sensors in a smartphone combined with its time and travel features, can identify when a user is intoxicated by cannabis with 90% accuracy. The study, which was published in the advanced online November issue of Drug and Alcohol Dependence, was done by Sang Wong Bae, an assistant professor at the Stevens Institute of Technology, who previously co-developed a smartphone app that detected binge drinking. To conduct the study, smartphone sensors to detect motion were monitored in young adults who reported that they use cannabis at least twice a week. More than 100 features were used to detect whether each participant was intoxicated, including GPS, noise, light, and activity levels. Researchers then looked at ‘day of the week’ and ‘time of day’ smartphone usage, while the participants reported when they were high or sober.

Related links

• Unbelievable moment drugged up driver flips car and smashes into payment booth on M6
• Dangerous levels of cocaine and MDMA found in Glastonbury river
• Katie Price 'admitted to police she was on drugs when they arrived at car crash'

Professor Bae and her colleagues, including those at Rutgers and Carnegie Mellon University, found that the combination of the two datasets predicted cannabis high with 90% accuracy in a natural environment. Bae then created artificial intelligence to detect a marijuana high, which can potentially be applied to detect risky behaviour, leading to early intervention in everyday settings. The Professor concluded: “It’s important to give people the chance to change their behaviour before something negative happens. "This study aims to predict human behaviour as a way to support people while physically or cognitively impaired.” Current marijuana detection methods such as blood, urine, or saliva tests have limitations. “This proof-of-concept study indicates the feasibility of using phone sensors to detect subjective cannabis intoxication in the natural environment, with potential implications for triggering just-in-time interventions," Bae said.